U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-286 241
MASS BALANCE DETERMINATIONS FOR POLLUTANTS IN URBAN
REGIONS. METHODOLOGY WITH APPLICATIONS TO LEAD, ZINC,
CADMIUM, AND ARSENIC
California Inst. of Technology, Pasadena
Prepared for
Environmental Monitoring and Support Lab. , Las Vegas, NV
August 1978
-------
&EPA
Environment! Mrunto
and Suppon I ;iUn;itur
PO Box ISO?/
Las Vegab NVS9Hd
EPA-600-4 78 040
August 1978
PB 286 241
Environmental
Monitoring Series
Mass Balance
Determinations for
Pollutants in
Urban Regions
Methodology with
Applications to Lead
Zinc, Cadmium, and Arsenic
REPRODUCED BY
NATIONAL TECHNICAL
INFORMATION SERVICE
U. S. DEPARTMENT OF COKMERCC
spmnmtui.vn.ai6i
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING ser.es.This series
describes research conducted to develop new or improved methods and instrumentation
for the identification and quantification of environmental pollutants at the lowest
conceivably significant concentrations. It also includes studies to determine the ambient
concentrations of pollutants in the environment and/or the variance of pollutants as a
function of time or meteorological factors.
This document is available to the public through the National Technical Informatron
Service. Springfield. Virginia 22161
REPORT NO.
EPA-600/4-78-046
TITLE AND SUBTITLE ^55 BALANCE DETERMINATIONS FOR
JOLLUTANTS IN URBAN REGIONS Methodology with
Applications to Lead, Zinc, Cadmium, and Arsenic
TECHNICAL REPORT DATA
(Please read Instruction* on the revene before completing)
[5. fl
August 1978
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Department of Environmental Health Engineering
PERFORMING ORGANIZATION NAME AND ADDRESS
California Institute of Technology
Pasadena, California 91125
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency-Las Vegas, NV
Office of Research and Development
Environmental Monitoring and Support Laboratory
Las Vegas, NV 89114
I. PERFORMING ORGANIZATION REPORT NO.
1HD620, 1HD622 '
11. CONTRACT/GRANT NO.
68-03-4034
13. TVPE OF REPORT AND PERIOD COVERED
Special Report
14. SPONSORING AGE
EPA/600/07
5.SUPPLEMENTARY NOTES ^^ pro.ject was partially funded by EPA's Technology Transfer
ffice under the Environmental Research Information Center, Cincinnati, Ohio.
6. ABSTRACT
A methodology is presented for constructing mass balances for pollutants which move
interactively through the air, land, and water of an urban-industrial region. Results
are reported for lead, zinc, cadmium, and arsenic based on experiments conducted
specifically for this study, and on available data from the open literature. The
principle on which the analysis is based is the conservation of mass equation for a
given chemical element. Using chemical element balance as in flow diagrams for the
movement of pollutants through the environment, rates of flow and accumulation can be
estimated for the separate environmental compartments.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
Pollutant inventories
Chemical balances
Particle size-distribution
Systems analysis
RELEASE TO PUBLIC
Mass balances
Pollutant pathways
Particle dispersion
Transport of lead, zinc,
cadmium and arsenic
19. SECURITY CLASS i
UNCLASSIFIED
20. SECURITY CLASS (TTilj pagtl
UNCLASSIFIED
COSATI Held/Group
13B
07B
21. NO. OF PAGES
126
EPA Fora 2220-1
-------
EPA-600M-78-046
August 1978
MASS BALANCE DETERMINATIONS FOR POLLUTANTS IN URBAN REGIONS
Methodology with Applications to
Lead, Zinc, Cadmium, and Arsenic
by
Department of Environmental Health Engineering
California Institute of Technology
Pasadena, California
Contract No. 68-03-4034
Project Officer
Edward A. Schuck
Monitoring Systems Research and Development Division
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and
Support Laboratory-Las Vegas, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents
necessarily reflect the vievs and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
ii
Foreword
Protection of the environment requires effective regulatory actions
which are based on sound technical and scientific information. This
information must include the quantitative description and linking of pollutant
sources, transport mechanisms, interactions, and resulting effects on man and
his environment. Because of the complexities involved, assessment of specific
pollutants in the environment requires a total systems approach which trans-
cends the media of air, water, and land. The Environmental Monitoring and
Support Laboratory-Las Vegas contributes to the formation and enhancement of
a sound monitoring data base for exposure assessment through programs designed
to:
develop and optimize systems and strategies for moni-
toring pollutants and their impact on the environment
• demonstrate new monitoring systems and technologies by
applying them to fulfill special monitoring needs of
the Agency's operating programs.
This report discusses a methodology for constructing mass balances
for pollutants which move interactively through the air, land, and water of
an urban- industrial region. The method was developed for application in
determining the fate of potentially toxic substances which may be conserved
in the environment. Such application will be useful to State and local
pollution control agencies in designing monitoring systems for determining
pollutant levels in various environmental compartments. The Monitoring
Systems Design and Analysis Staff at this Laboratory may be contacted for
further information on the subject.
George B1. Morgan
Director
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada
ill
-------
PREFACE
This report was edited by personnel at the U.S. Environmental
Protection Agency's Environmental Monitoring and Support Laboratory, Las
Vegas (EMSL-LV) from a final report submitted to the EMSL-LV by the
Department of Environmental Health Engineering at the California Institute
of Technology (Caltech). The Galtech final report summarized work-to-date
on area-wide naas balances techniques and Included several papers which were
published in the open literature that were authored by staff members of the
Department of Environmental Health Engineering. Different portions of the
original manuscript were written by C. I. Davidson, S. K. Friedlander,
J. J. Huntaicker, R.-C. Y. Koh, J. J. Morgan, J. Vucela, and N. H. Brooks..
iv
CONTENTS
Foreword Hi
Preface iv
Figures vi
Tables > viii
Abbreviations ix
I. Introduction 1
II. Summary 3
Mass Balances for Trace Metals: State-of-the-Art . . 8
Mass Balances for Trace Metals: Future Prospects . . 11
III. Relation Between Airborne Concentration and Deposition
of Trace Metals 15
Theory of Deposition 18
Size Distributions 23
Atmospheric Size-Distributions: Isokinetic
Experiments 27
Deposition of Lead, Zinc, and Cadmium ..,..,. 32
Experimental Errors 37
Andersen Impactor Cutoff Diameters 40
Discussion of Trace-Metals Deposition Studies. ... 42
IV. Examples of Mass Balances 44
Automotive Lead in the Los Angeles Basin 44
Lead in the Saline Branch Watershed of Illinois. . . -70
Atmospheric Trace-Metal Flows in the Los Angeles
Basin: Zinc and Cadmium 71
V. Transport and Deposition through the Water Route 91
Flows of Arsenic in the Southern California
Water Environment 91
The Ultimate Fate of Sewage Farticulate Matter ... 96
VI. Concluding Remarks 105
VII. References 106
-------
FIGURES
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Example of a flow diagram for automobile-emitted trace
Airborne concentration of condensations nuclei as a
Data from figure 2, plotted on a logarithmic scale
Total deposition of lead on Teflon plates at various
heights.
Normalized size distributions for lead, zinc and cadmium,
based on aerodynamic equivalent diameters
Impactor stage collection efficiency as a function of
particle diameter dp, for a typical stage
Cumulative mass distributions for lead aerosol in auto
exhaust and in ambient air at Pasadena
Differential mass distributions for lead at 1-meter from
a freeway and for Pasadena
Distribution of deposition factors in the Los .Angeles
Basin
Deposition of lead as a function of distance from the
coast
The flow of automobile-emitted lead through the Los Angeles
Basin
Differential mass distributions for zinc at 1-meter from
a freeway and for Pasadena
Deposition as a function of distance from a freeway for
Pb, Zn, and Cd
Zinc fluxes in the Los Angeles Basin
Deposition of Pb, Zn, and Cd as a function of distance from
the coast
Page
6
20
21
22
30
41
52
53
55
60
68
77
79
81
85
vi
FIGURES (Continued)
Number
16
17
Vertical distribution of particle concentrations at the
centerline as a function of travel time ........
Measured and computer synthesized ocean current data off
White's Point, California .............
Page
. 100
. 103
vii
-------
TABLES
Number Page
1 Types of Information Necessary for Determining Fate
of Airborne Trace-Metal Pollutants 3
2 Percentage of Mass Greater Than 10 ym Aerodynamic Diameter
for Nonisokinetic and Isokinetlc Runs 28
3 Deposition Velocities of Lead 32
4 Calculated and Measured Fluxes.to Flat Surfaces 35
5 Airborne Concentrations During the Two Isokinetlc Impactor
Experiments 35
6 Values of Mass Mean Cutoff Diameters dp and the Standard
' Deviations for the Andersen Impactor Efficiency Curves. . 41
7 Pb/CO at Various Sites in the Los Angeles Basin 58
8 Input of Anthropogenic Los Angeles Lead to the Coastal
Waters 64
9 Mass Balance for Automobile-Emitted Lead 69
10 Zinc Content in Particulate Emissions from Metallurgical
Operations 73
11 Zinc to CO and Cadmium to CO Ratios in the Los Angeles
Basin 76
12 Mass Balance for Zinc in the Atmosphere 80
13 Mass Balance for Cadmium in the Atmosphere 84
14 Zn/Pb at Various Locations In the Los Angeles Area 86
15 Deposition Ratios, August 1973. . 87 .
16 Anthropogenic Inputs of Los Angeles Pb, Zri and Cd to the
Coastal Waters 89
17 Probability Definitions 102
vlii
ABBREVIATIONS
ACHEX
Caltech
cm/sec
CNC
dp
8/1
8/cm3
GDE
km
m/sec
MMED
ym
pg/m
ng/cra
nm
pdf
ppm
SCCWRP
u*
Vg
California Aerosol Characterization Experiment
California Institute of Technology
centimeters per second
condensation nuclei counter
aerodynamic equivalent diameter
median aerodynamic equivalent diameter
mass mean cutoff diameters
grams per liter
grams per cubic centimeter
General Dynamic Equation
kilometers
meters per second
millimeter
mass median equivalent diameter
micrometer and micron
microgram per cubic meter
nanograms per square centimeter
nanometer
probability density function
parts per million
Southern California Coastal Waters Research Project
friction velocity
deposition velocity
sedimentation velocity
Abbreviations are listed as used in this report and are not necessarily
the recognized standard abbreviations. For example, micron and micrometer
are used Interchangeably in the text, paraphrasing original input data
from the literature.
-------
SECTION I
INTRODUCTION
Studies of pollutant: flows through urban and industrial regions are
potentially useful in a number of ways. They can help in the design of
integrated monitoring systems for determining pollutant levels in various
environmental compartments and for determining control systems for limit-
ing pollutant exposures to prescribed levels. They can also be used to
estimate the effects of urban areas on surrounding regions. This is of
interest in the development of a rational nondegradation policy. Finally,
such studies are needed to determine the fate of potentially toxic sub-
stances, such as certain trace metals, which are conserved in the environ-
ment.
This report is a review of the methodology which has been developed
at the California Institute of Technology (Caltech) for constructing mass
balances for pollutants which move in a connected manner through the air,
land, and water of an urban-industrial region. Results are reported for
lead, zinc, cadmium, and arsenic which are of special interest to the
U. S. Environmental Protection Agency (EPA) and for which a reasonable
amount of data were available. A balance has also been constructed for
sulfur and will be described in a subsequent report. The principle on
which the analysis is based is the conservation of mass for a given
chemical element. Using chemical element balances in flow diagrams of
the movement of pollutants through the environment, flow rates and
accumulation can be estimated for the separate environmental compartments.
This method of constructing mass balances was first developed and
applied to lead, zinc, cadmium, and nickel under a grant from the Rockefeller
-------
Foundation to Caltecb and published in 1975 (Huntzicker and Davidson,
1975). The mass balance for lead was more completely detailed than that
of zinc and cadmium. The data for nickel were not sufficient to allow
any conclusions to be made. In the Huntzicker and Davidson (1975) report,
the concept of near and far deposition was introduced and the various
environmental pathways through the air, land, and water of the Los Angeles
area were discussed. However, little was known about the physical mech-
anisms of trace-metal transport in the various environmental compartments.
Advances have been made since the first mass balances report was
published. The most significant gain involves new data on the physical
mechanisms of trace-metal deposition. The deposition of particles
containing trace metals far from the source can be predicted fairly
accurately based on an understanding of these fundamentals. This method
of predicting the far deposition of lead, zinc, and cadmium is described
in Section III. Application of the advective diffusion equation to describe
particle transport in the ocean has also been achieved, and this is covered
in Section V.
Because of uncertainties in this study, the mass balance for nickel has
not been Included in this report. The flow of arsenic through the water
pathways of the Los Angeles Basin, however, has been investigated and is
presented in Section V.
SECTION II
SUMMARY
The development of a mass balance can be carried out by using a flow
diagram for each species of interest. The data necessary to prepare such
a diagram can be divided into three main categories: source or input data,
receptor site data, and output data. The last category refers to the flow
of the pollutant out of the urban area and into adjoining areas. The types
of information required for each category are listed in Table 1.
TABLE 1. TYPES OF INFORMATION NECESSARY FOR DETERMINING
FATE OF AIRBORNE TRACE-METAL POLLUTANTS
Source or Input Data
Identity of sources
Mass emission rate for each source
Aerosol size distribution for each source
Receptor Site Data
Deposition rate (dry)
Rainout-washout rate
Output Data
Atmospheric flushing (ventilaton) rates (i.e., inputs to contiguous,
nonurban regions)
Receptor site atmospheric concentration
Volumetric flow through the air basin.or appropriate tracer
(Continued)
-------
TABLE 1. TYPES OF INFORMATION NECESSARY FOR DETERMINING
FATE OF AIRBOKSE TRACE-METAL POLLUTANTS (Continued)
Inputs to coastal waters, lakes, rivers, etc.
Deposition rate (dry)
Rainout-washout rate
Runoff
Sewage
The primary criterion for success in an analysis of pollutant flow
is that a satisfactory mass balance be obtained; that is, the sum of the
input flows should equal the sum of the output flows and stored quantities.
However, such an analysis is not possible without the source emission rates.
The lack of such rates is the primary difficulty in this type of analysis.
An inaccurate emission Inventory can sometimes result in the failure to
identify an important environmental pathway.
The major environmental pathways for an aerosol are by ventilation (re-
moval from the urban area by wind) and deposition within the urban area. The
ventilation rate is estimated by assuming that the air basin is a continuously
stirred chemical reactor. In this approximation the ventilation rate., q., of
species 1 is:
[i] Q
(1)
where [1] is the average receptor site concentration of species 1 and Q is
the volumetric air flow through the basin. In general, however, Q is not
known, and a tracer method must be used. One such tracer for the volumetric
air flow is carbon monoxide which is essentially an unreactive species.
Because of its nonreactivity the input and outflows of carbon monoxide are
equal. With carbon monoxide as a tracer, the following approximate relation-
ship was used to determine q, .
q =
(2)
[CO] is the atmospheric concentration of carbon monoxide and q« is the
known Input rate of carbon monoxide. This approach works particularly well
when carbon monoxide and species i have a common source (e.g., Pb and CO)
and less well when the sources are different.
An example of the flow of automobile-emitted lead through the Los Angeles
Basin is represented in a general flow diagram for trace metals emitted by
the automobile shown in Figure 1. This diagram shows the transport of a
pollutant from the source and its eventual fate in the environment. Three
main possibilities are indicated. Some material may become bound to the soil
after deposition. Other material may deposit in local bodies of water,
where it may remain in solution or suspension, or sink to the sediments.
Finally, the pollutant may remain airborne and leave the urban region. The
particular pathway followed, as well as the eventual fate of the material,
is of great importance in terms of environmental effect and chemical
speciatlon. In Figure 1, physical processes are designed by rectangles
and receptors are represented by ellipses. Emissions from an air pollution
source generally include a wide variety of particles and gases. Most of the
trace metals, with which this report is concerned, are associated with
the particulate component.
The greatest loss rate of particles from the atmosphere occurs
nearest the source. In the case of automobile emissions, large particles,
such as reentrained tailpipe material and tire dust, may deposit in
relatively large quantities on or near roadways. The rate of deposition
usually decreases with Increasing distance from the source.
Environmental effects of near and far deposition are likely to be
different. Near deposition occurs mainly on the roadway and, therefore,
significant fractions of material ultimately reach sanitary landfills, as
well as the coastal waters. Most of the landfill material becomes
immobilized, while the trace metals reaching the coastal waters enter as
point sources.
-------
Figure 1. Example of a flow diagram for automobile-emitted trace metals.
6
On the other hand, far deposition occurs over large regions, acting
as an area source rather than as a point source to the coastal waters.
While near deposition involves "hot spots" of material, far deposition
is more evenly spread over a large area.
Near deposition is controlled by gravitational sedimentation. Because
of the short residence tines of large particles in the atmosphere, meteo-
rology has little influence on near deposition. Surface roughness is also
relatively unimportant, except when the height of the roughness elements
becomes large enough to cause inertlal impaction.
Far deposition of trace materials has been shown by recent data to be
controlled by sedimentation for smooth, flat surfaces and low windspeeds.
In the case of rough surfaces and moderate windspeeds, however, it is likely
that inertial impaction is important. Meteorological effects strongly
influence far deposition.
Figure 1 traces the pathways followed by particles through the
environment. Note that plants and animals, which may be harmed by the
particles, are receptors in a number of different pathways. Near depo-
sition may occur on streets or on the soil and vegetation. Traffic on
a road located near a body of water may be a source for direct deposit
on the water. However, particles are more likely to reach bodies of
water by being carried off the streets via surface runoff. Street
sweeping is also important since road dust is brought to sanitary land-
fills where it is buried in the soil. This is discussed in greater
detail in Section IV.
Once particles reach the soil by either near or far deposition,
there are three possible routes which may be taken. If the particles
become bound to soil particles, they may become immobilized. If the
particles are soluble and the soil porous, they may be carried by rainwater
percolation into the groundwater system. If the particles have deposited
on the surface of the soil, they may remain there until storm runoff
washes them away; subsequently dissolved metals may be artificially
-------
recharged into local groundwater basins. These possibilities are discussed
further in Section IV. In either of the latter two routes, the material
will eventually reach a body of water. Large particles will then sink to
the sediments, while smaller particles may dissolve and remain in solution.
The material may be consumed by plants or animals, becoming part of the
food chain before eventually dissolving in solution or sinking to the
sediments.
MASS BALANCES FOR TRACE METALS: STATE-OF-THE-ART
The most complete analyses made to date have been carried out for
lead in the Los Angeles Basin. The results of these analyses can be
summarized in the form of a matrix equation as follows:
[P]
(3)
where S • sink vector [S. S, S, ... S ]
1 i j n
E = source emissions vector [E. E- E, ... E ]
123 m
[P] = source-to-sink matrix, n x m
Here the elements of S and E have units of metric tons* of lead per
day,•while the elements of [P] are the fractions of lead emitted from each
source ultimately reaching each sink. The sum of elements in the first
column [P] is not unity as expected under ideal conditions since the
agreement between input and output routes was not perfect.
The source and sink vectors are as follows:
S. = retained in vehicle
S2 = buried in sanitary landfill
S_ = deposited on soil, vegetation
s"referes
S, " remains airborne
Sq = enters coastal waters
E. = vehicle emissions =23.7
E- = sewage =0.6
- -
sl
S2
S3
S4
S5
r -i
|0. 245 0.0
JO. 060 0.0
= jO.410 0.0
|o.220 0.0
JO. 035 1.0
El
E2
So, S
5.8
1.4
9.7
5.2
1.4
(4)
The total lend mass balance is developed and presented in Section IV.
It is shown that fair agreement has been obtained between input and output
quantities indicating that most of the major sources and sinks have been
accounted for.
Lead is the only element, however, for which a relatively complete
mass balance has been obtained. One reason for this is that an accurate
source emission inventory has been used. Lead emissions can be readily
estimated since there is only one major source category. Data on fuel
consumption and the lead content of gasoline can be obtained, hence
calculating lead emissions is relatively simple. Further the levels of
lead found in the environment are well within the range of sensitivity of
current analytical techniques, thus simplifying sampling and analysis
and assuring that accurate data can be obtained.
Although a reasonable lead mass balance has been developed, a complete
picture is not available. The eventual fate of lead buried in sanitary
landfills is still unknown, as is the fate of lead carried out of the Los
Angeles Basin by winds is uncertain. Also, the amount of lead removed from
the atmosphere by precipitation needs further study.
-------
The construction of a mass balance for other elements is more
difficult than for lead. The major obstacle in the construction of a
zinc balance is the difficulty in obtaining accurate source emissions data.
Eclifce lead, which is emitted from one major type of source, zinc comes from
* variety of mobile and stationary sources.
Zinc emissions from mobile sources result from the abrasion of truck
v-A automobile tires, and from combustion of detergent lubricating oils.
The zinc content of tire rubber and lubrication oils has been estimated,
aitd the zinc emissions rates have been calculated in Section IV. However,
there are large uncertainties in these calculations. For example, zinc
deposition measurements near a Los Angeles freeway could account for only
10 percent of the calculated tire dust zinc emissions. This suggests
that tire dust deposits directly on the roadway and in areas where tire
abrasion is more severe.
It.is also difficult to obtain accurate estimates of zinc emissions
from stationary sources. These are mainly metallurgical operations. The
method of obtaining stationary source emission estimates shown in Section
IV involves using data on the total mass of particulate emissions along
with estimates of the zinc content of such emissions. The percentage of
zinc varies greatly from plant to plant, as does the total mass of
particulate emissions. Hence the zinc emissions from stationary sources as
well as mobile sources can only be roughly estimated. As a result, only
an approximate mass balance has been obtained for zinc. The fair agreement
between input and output routes suggests that the major sources and sinks
have been Included in spite of the uncertainties.
A satisfactory mass balance has not been obtained for cadmium. Data
compiled in Section IV show that the outputs greatly exceed the inputs. This
suggests that all of the sources of cadmium have not been accounted for and
it is likely that both the mobile and the stationary source emissions
estimates are low. For example, measurements of cadmium could, account for
only 70 percent of the calculated tire dust cadmium emissions. Stationary
source cadmium emissions are likewise difficult to estimate. There are a
10
number of metallurgical operations and cadmium chemical plants in the area,
but little data are available on emissions from these sources.
In addition to difficulties in obtaining source data, the levels of
cadmium present in the Los Angeles area are very low. Analyses are difficult,
and background levels are high. More research is needed to obtain a reason-
able mass balance for cadmium.
Very little data on arsenic are available, and hence no mass balance was
constructed. However, it is likely that there may be no major sources of
atmospheric arsenic in the Los Angeles area. Phosphate plants in the Los
Angeles Basin may be a source of certain trace metals such as zinc, cadmium,
and arsenic but emissions from these phosphate plants were not considered
in the present study. Thus, only the water route for arsenic was examined
and this is discussed in Section V.
In summary, fair agreement between input and output masses has been
obtained for a few elements using the guidelines set up in this report. It
should be possible to gather sufficient data to construct mass balances for
additional elements of public health and ecological interest by this method.
Obtaining accurate source emissions data, identifying environmental path-
ways and sinks, and determining the source-to-sink relationships are the
keys to constructing accurate balances. Insufficient data exist for some
toxic elements of interest and research in these areas should be encouraged.
MASS BALANCES FOR TRACE METALS: FUTURE PROSPECTS
In the previous discussion it has been assumed that the Los Angeles
Basin is a continuously stirred chemical reactor because it is a relatively
isolated meteorological area in which reactions are occurring. Future mass
balance studies are likely to be based on more detailed methods permitting
concentration estimates of more widely dispersed agents of interest.
For example, from a fundamental point of view, the transport and
fate of particulate matter emitted by the automobile or found in stack plumes
11
-------
can be estiaated by making use of the General Dynamic Equation (GDE) for
determining aerosol behavior (Friedlander, 1976). This equation takes Into
account particle formation, growth, diffusion, coagulation, deposition,
and advection by the winds. The GDE for an external force field is shown
as equation 5. Here n(v) is the number distribution function defined for
all particle volumes v, and / n(v) dv • N, the total number concentration
of particles in all size ranges.
-/.
^ / B (v, v-v)n(v)n(v-v) dv
B (v,v)n(v)n(v)dv - 7-cn (5)
time
v = velocity of the fluid suspending the particle
D = Brownian diffusion coefficient
-*•
c = particle velocity relative to fluid due to an external force field
s - particle growth law dv/dt
ia
The first term 3t represents the time rate of change of the function n
due to the various mechanisms. This term must be nonzero unless the system
is in a steady state.
The second term V-nv represents advection of particles, for example,
by wind. For the usual case of incompressible flow, V-v is zero and this "
term becomes v»7n.
The third term takes particle growth into account. Growth here refers
to any gas-to-particle conversion process.
On the right hand side V'DVn represents Brownian dlffuslonal transport.
Since in general D is not a function of position, this term becomes DV n.
12
The second term on the right hand side describes the formation rate of
r\j
particles of size v due to the collision of smaller particles of size v-v
r\f
and v. The term following also represents coagulation which in this case
is the loss rate of particles of size v with all other sizes.
Finally, the last term on the right hand side represents particle
motion in the influence of an external force field. In the case of
gravitational sedimentation, V-cn merely becomes v -5—, where v is the
settling velocity which is taken to be negative in the -z direction.
For the atmosphere, a form of the GDE for turbulent conditions must
be used. It is assumed that the fluid velocity v, the number distribution
function n, and the particle growth law s can be written as the sum of the
mean and fluctuating components:
-•---
V = V + V
n = n + n' (6)
s = s + s'
Substituting these relations into equation 5 and averaging with respect
to time yields the following result for steady turbulent flow:
~
b — — j
. v-v)n(v)n(v-v)dv
B(v,v)n(v)n(v)dv + j f B(v,v-v)n' (v)n' (v-v)dv
Jo
B(v,v)n'(v)m'(v)dv - v
(7)
13
-------
The third term on the left hand side is the fluctuating growth term
which depends on the correlation between n' and the local concentration of
gaseous species converted to particulate matter. The first term on the
right hand side represents the change in n resulting from turbulent
diffusion; the individual components of the vector flux n'v' are usually
assumed to follow an equation of the form:
1 i 3xt (8)
where the eddy diffusivity e± is a function of position. The fifth and
sixth terms on the right hand side are the contributions to coagulation
resulting from the fluctuating concentrations.
Although the GDE has not yet been used to construct mass balances,
studies related to this* equation have been conducted for future mass
balance calculations. Studies discussed in Section III on the physical
mechanisms of far deposition for lead, zinc, and cadmium in the Los
Angeles Basin involve the determination of the deposition velocities of
these species. Such data will be of value in the development of deposition
models based on the GDE. Numerical solutions to the advective diffusion
equation applied to sewage particles in the ocean are examined in Section V.
This form of the GDE may have direct application to mass balances but this
work has not yet been conducted.
SECTION III
RELATION BETWEEN AIRBORNE CONCENTRATION AND DEPOSITION OF TRACE METALS
The deposition of trace metals is expected to decrease with distance
from the source, i.e., a roadway. Most of the decrease occurs within about
100 meters from the road; at greater distances the deposition is
essentially independent of distance and is termed "far deposition." The
"near deposition" region has been defined as the distance over which the
deposition rate falls to within 10 percent of the far deposition rate.
It is understood that the far deposition rate may vary in different
parts of an urban area. Thus the size of the near and far deposition
regions as well as the actual deposition rates may also vary within the
area.
Although the general dynamic equation cannot be solved for most cases
involving atmospheric transport, there are a few situations where the
basic theory may be directly applied. One such case Is the deposition
of certain particles containing trace metals on smooth, flat surfaces.
The deposition of particles containing lead, zinc, and cadmium on surfaces
such as smooth soil, roadways, and other flat areas may be directly
predicted under certain conditions. In particular, these predictions may
be made when there is sufficient mass in large partlculates so that
gravitational sedimentation is important, and when the windspeed is not
so excessively high that turbulent deposition dominates.
Airborne concentrations of trace metals In urban areas have been
measured by several investigators, and it is generally known that urban
15
-------
air contain* significantly higher trace metal levels than air from
nonurban regions. Lee and von Lehmden (1973) have shown that urban
activities such as fuel combustion contribute a major portion of the trace
cetals emitted into the atmosphere. Their data show concentrations of iron,
lead, magnesium, and zinc in excess of 1 microgram per cubic meter (pg/m )
in many cities. Lazrus et al. (1970) have examined trace metals in
precipitation throughout the United States and have found considerably
greater concentrations near population centers. For example, they found
the lead content of rainwater to be correlated with gasoline consumption
in a given area.
Airborne concentrations of lead have been measured near roadways by
many investigators (Bullock and Lewis, 1968; Cahill and Feeney, 1973;
Daines et al., 1970; Rolfe and Haney, 1975). Studies by Cardina (1974)
and Pierson and Brachaczek (1974) have shown that particles of rubber tires
can be identified in the dust collected near roadways. This dust contains
significant amounts of zinc which contributes to the total airborne zinc
concentration in urban areas. Metallurgical emissions also contribute to
urban levels of zinc and other trace metals.
A knowledge of deposition rates for these trace metal particles has
many applications. Information on particle residence time in the atmosphere
can be obtained from deposition studies. The effects of sources on down-
wind regions are linked to deposition rates. Deposition rates are also
useful when setting up tracer studies.
Most trace-metal deposition studies have focused on lead-containing
particles. Rablnowitz (1972) has measured the lead content of wild oat
grass (Avena eativa) in urban and nonurban areas of Southern California.
The lead associated with these plants was shown to be deposited airborne
material rather than internally incorporated material from the soil.
Patterson and Settle (1974), Huntzicker et al (1975a), and Davidson et al. .
(1974) have examined the deposition of lead on flat Teflon plates which were
placed in cities as well as in the mountain, desert, and coastal areas of
16
Southern California. Similar studies involving the lead deposition on
artificial surfaces have been conducted by Atkins (1969) in Palo Alto and
by Servant (1974) in France. The results of several lead deposition
studies are summarized in a report by Chamberlain (1974).
The specific case of automobile-emitted lead depositing on soil and
vegetation near roadways has been examined by several investigators. Kloke
and Riebartsch (1964) have measured a high lead content in grass growing
near busy streets. Heichel and Hankin (1972) have detected large lead
particles deposited on trees near a roadway. The lead content of soil '
near a highway was compared with lead in soil samples obtained near a
battery smelter and from a greenhouse soil supply (Marten and Hammond,
1966). Rapid decreases in the lead content of vegetation with increasing
distance from highways have been measured by several experimenters
(Rolfe and Haney, 1975; Cannon and Bowles, 1962; Cholak et al., 1968;
Chow, 1970a; Page et al., 1971). Other researchers have examined several
varieties of plants with high lead contents growing near roads, investi-
gating whether the lead is deposited airborne material or translocated
from the soil (Dedolph et al., 1970; Leh, 1966; Motto et al., 1970; Schuck
and Locke, 1970; Ter Haar, 1970; Rabinowitz, 1972).
The deposition of other trace metals has also been examined. Peirson
et al. (1973) have studied the deposition of several elements in England.
Lagerwerff and Specht (1970) have measured the lead, zinc, nickel, and
cadmium content of soil and grass at various distances from roadways in
Eastern and Midwestern United States. They found a decrease in metal
content with increasing distance from the road. Similar gradients have
also been measured by Huntzicker and Davidson (1975) near a Los Angeles
freeway. In other studies, Huntzicker et al. (1974) have measured the
deposition of zinc, nickel, and cadmium on flat Teflon plates in Southern
California.
Although there is a wealth of trace-metal data available, little
research has been done to investigate the deposition mechanisms involved.
Several investigators have measured the deposition of monodispersed
17
-------
particles la Che laboratory and under controlled field conditions, but these
results have not previously been applied to ambient trace-metal deposition.
The work discussed in this report involves experimental determination of
airborne size distributions of particles containing lead, zinc, and cadmium
in Pasadena. The distributions are then used to calculate deposition on a
flat surface. The calculated values are compared with experimental
deposition data obtained simultaneously with the size-distribution measure-
ments. It is expected that these calculated and measured depositions for
smooth, flat surfaces will be the lower limits of the actual deposition
on rough, natural surfaces.
All particle diameters refer to aerodynamic equivalent diameters,
unless otherwise noted. A particle with an aerodynamic equivalent
diameter (dp), regardless of shape or density, will have the same inertial
properties as a unit-density sphere of diameter dp. The value of dp in a
size-distribution of particles where 50 percent of the total mass is greater
and 50 percent smaller than this diameter is defined as the mass median
equivalent diameter (MMED). The geometric diameter is the actual distance
across the particle; it is thus difficult to apply this concept to an
irregularly shaped particle.
THEORY OF DEPOSITION
It is convenient to begin by defining the deposition velocity v :
V8 • -C
(9)
where v has units of centimeters per second (cm/sec), J is the flux of
material to a surface in nanograms per square centimeters per second
(ng/cm /sec), and C is the airborne concentration in nanograms per cubic
centimeter (ng/cm ). The flux is assumed to be constant with height above.
the surface, although C and v are functions of height.
18
The relationship between flux and concentration can be written:
(IH-E) - vgC (10)
where D and e are the Brownian and eddy dif fusivltles, respectively, in
2
square centimeters per second (cm /sec), z is the height above the surface
in cm, and v is the sedimentation velocity in cm/sec. The sedimentation
s
velocity of a particle is reached when the aerodynamic drag force on a
falling particle exactly balances the gravitational force. The minus signs
are necessary since the flux downward to a surface is a negative quantity by
convention and the deposition velocity is defined as positive.
Equation 10 applies to both small and large particles. For very
small particles, the sedimentation term is negligible. It is also assumed
that D is very much less than e, except very close to the surface or when
there is no wind. Thus for small particles, introducing the relationship
e = ku*z, equation 10 reduces to:
dC
dC . . dC
J = -e -T— = -ku*z -7—
dz dz
-ku*
d(ln z)
(ID
where k is von Karman's constant equal to 0.4, u* is the friction velocity
in cm/sec, and In z is the natural logarithm of height.
For large particles, the diffusivities are negligible so equation 10
becomes:
J = -v C (12)
Comparing equations 9 and 12 shows that the deposition velocity equals
the sedimentation velocity for large particles.
Equation 11 may be used to calculate the deposition of small particles
when dC/d(ln z) is known. For example, Figure 2 shows a plot of condensa-
tion nuclei concentration versus height over a field of wild oat grass In
residential Los Angeles. The value of C(4 meters) is 45,000 particles
per cubic centimeter. Two condensation nuclei counters (CNC) were used
19
-------
to make these measurements (General Electric CNC, and Environment One Model
Rich 100). A hot wire anemometer was used to determine u*. When plotted
on a logarithmic scale of height, as ID Figure. 3, a value of dC/d(ln z) is
obtained which may be substituted into equation 11. The resulting
2
calculated flux for these experiments is -40,000 particles/cm /sec. The
deposition velocity from equation 9 is thus 0.9 cm/sec at a height of 4
meters.
400
§
o 300
ZOO
S 100
u'=43 em/SBC.
— HEIGHT OF VEGETATION
.20 .40 .60 .80
C(Z)/C (4 METERS)
1.00
Figure 2. Airborne concentrations of condensation nuclei as a
function of height.
The linear relationship shown in Figure 3 implies that the condensation
nuclei were approximately monodisperse. If the nuclei size-distribution
were polydisperse, the different-sized particles would have correspondingly
different fluxes, hence different dC/d(ln z) values. Therefore the overall
effective dC/d(ln z) would probably vary with height.
20
500
E 200
a 100
| 50
is
jj
5 20
.20 .40 .60 .(
C(Z)/C (4 METERS)
1.00
Figure 3. Data from Figure 2 plotted on a logarithmic scale.
In the case of large particles, a different concentration profile is
expected. Figure 4 shows a plot of lead mass-deposition on flat Teflon
plates. The plates were located at several heights above the roof of the
U. M. Keck Laboratories at Caltech about IS meters above the ground. If the
lead particles depositing on each plate are from the same portion of the
airborne size-distribution, then the deposited mass is proportional to the
airborne concentration of lead in that size range.
21
-------
350
300
! 200
I
i iso
i 100
so
10 20 30 40 SO
Ft DEPOSITION, nj/cnj' OAT
Figure 4. Total deposition of lead on Teflon plates "at various heights.
At heights above 1 meter, the deposition is roughly constant with
height. Below this height, the deposition increases, indicating a source
of lead particles on the roof surface. This is probably due to resuspension
of roof dust by wind eddies. This premise is supported by a measurement
of high lead content in the roof dust in a subsequent experiment.
It should be mentioned that although equation 10 simplifies to
equation 11 or 12 for small or large particles, the general case has
been examined by other researchers. For example, equation 10 has been
integrated by Chamberlain (1966) and shown in greater detail by Sehmel (1970),
based on special assumptions concerning the form of a e near the surface.
22
It will be shown that equation 12 is applicable for certain trace
metals at the CalTech site because there is sufficient mass in large
particles to control deposition by sedimentation.
SIZE-DISTRIBUTIONS
Many investigators have measured lead size-distributions at the source,
the automobile tailpipe. Hirschler et al. (1957, 1964) used an electro-
static precipitator to collect auto exhaust particles. They then washed the
particles into a solution and fractionated them by repeated centrifugal
sedimentations. The results of their experiments showed that particles
containing lead in auto exhaust covered a wide spectrum of sizes, from
0.01 micrometers (ym) to several millimeters in diameter. For city-type
driving, one-third to one-half of the lead was in particles greater than
5 ym in geometric diameter (12.2 ym aerodynamic equivalent diameter) assum-
ing that lead bromochloride (PbBrCl) has a density of about 6 grams per
cubic centimeter (g/cm ). Nlnomiya et al. (1971) measured the sizes of
particles containing lead in diluted auto exhaust with vehicles operating
under the 7-oode cycle of the Federal Test Procedure (Federal Register
No. 108, 1968). Their results showed 20 to 30 percent of the exhausted lead
particles were of 500 to 5,000 ym in geometric diameter and were expelled in
the first fev cycles after a cold start. Ter Haar et al. (1972) has used
the Federal Test Procedure and has collected auto exhaust in a large black
polyethylene bag. Sizing these particles showed that 55 percent of the lead
mass exhausted was associated with particles greater than 5 ym in.equivalent
diameter. It is interesting to note that Ter Haar found a mass median
equivalent diameter (MMED) of only 0.5 ym when the vehicle was operated under
steady cruising conditions rather than the Federal cycle. Habibi (1973) has
reported that 57 percent of the particles containing lead in auto exhaust
are greater than 9 ym equivalent diameter, for a car with 28,000 miles
(45,000 km). . His vehicles were operated under the Federal Mileage
Accumulation Schedule (Federal Register No. 2, 1968).
23
-------
Two Investigators have conducted similar tailpipe studies but have not
seen significant lead In large particles. Mueller et-at.-(1964) found 62 to -
80 percent of the lead mass In particles smaller than 2 pm In equivalent
diameter. His studies Involved automobiles operated under steady cruising
conditions, and the results are consistent with Ter Haar's lead MHED of
0.5 Urn for steady cruising. Lee et al. (1971) operated an automobile
according to a 5-mimite cycle similar to the California cycle. The exhaust
was sampled from animal inhalation exposure chambers. These studies showed
that 95 percent of the lead mass was in particles smaller than 0.5 Urn in
equivalent-diameter. -
Difficulties in performing these tailpipe experiments as well as
variations in automobiles, fuel, and driving cycles account for the
differences among the results of the above Investigators. Hablbi (1973)
discusses some of these difficulties. In general, however, the results
Indicate that relatively new, steadily crusing automobiles emit predominately
submicron lead. Older autos and rapidly accelerating or decelerating autos
emit both micron and submiron lead particles. It is therefore likely that
ambient urban atmospheres which receive emissions from wide varieties
of vehicles and driving patterns contain a broad spectrum of sizes of
particles containing lead.
The situation for particles containing zinc and cadmium has also been
investigated. Metallurgical emissions and automobiles (tire dust) are
the major sources of zinc. The metallurgical emissions are generally
submicron while the tire dust is larger. It is therefore likely that much
of the zinc deposited in urban areas, especially near roadways, is from
tire dust.
Cardina (1974) has measured the size distribution of road dust a few
meters from an urban freeway. He found one-third of the total mass in
particles larger than 7 Mm in equivalent diameter. These large particles
had a high mass fraction of rubber, presumably from tire dust, and would
therefore contain a significant amount of zinc.
24
Pierson and Brachaczek (1974) have measured both airborne and settled
highway dust. Their data show that most of the lead and zinc deposited
near roadways are large particles (50 to 500 pan in geometric diameter), and
that most of the tire dust is found in large, nearly nonsuspendlble particles
of which only a small fraction becomes airborne. Dannis (1974b) has
examined tire dust collected on plates mounted underneath a car. His
findings show a MMED for rubber of about 20 pm, with very few rubber
particles smaller than 3 vm in equivalent diameter.
These investigations indicate that ambient particles containing lead
should be widely distributed over a large range of sizes. The ambient zinc
distribution should be bimodal: a submicron peak from combustion and
metallurgical emissions, and a supermicron peak from tire dust. The ambient
cadmium distribution may resemble the zinc spectrum. Substances which
contain zinc generally contain trace amounts of cadmium because the
two elements are found together in nature (Friberg et al., 1974). In
some regions, additional sources of cadmium may be present.
Measurements of ambient trace-metal size-distributions have been
concerned with the respirable range near and below one micron. These
studies were generally nonisokinetic and therefore underestimated the
amount of material in large particles. Watson (1954) has shown that
sampling with an air inlet oriented 90° to the wind, such as an upright
impactor, results in collecting less than 45 percent of the 12 vm particles
and less than 10 percent of the 37 ym particles present (equivalent
diameters).
Tufts (1959) measured sizes of particles containing lead using a
chemical microscopic spot test, sizing red dots of precipitate. The
results for ambient air near heavily traveled streets showed the bulk of
the lead in particles between 0.1 and 2.7 pm in geometric diameter. There
was lead detected in particles as large as 13.4 Mm. Lininger et al. (1966)
measured airborne lead near a busy street and found large variations in
the size-distributions among runs. Robinson and Ludwig (1967) used a
25
-------
r-odlfloi Goetz spectrometer (Goetz et al., 1967) to obtain size-distributions
of lead in remote and urban areas throughout the United States. Surprisingly,
little variation in the shapes of the distributions was found, with the
average MMED being 0.25 van. Cholak et al. (1968) measured lead size-
distributions 10, 400, and 3,400 meters from a busy interstate highway with
Andersec impactors. About 70 percent of the lead mass was found in
particles smaller than 1 |jm in equivalent diameter. The extrapolated MMED
was about 0.3 ]m. Negligible differences appeared among the distributions
from the three sites. Daines et al. (1970) used cascade impactors located
10, 75, and 500 meters from the highway. They found 65 percent of the lead
mass in particles smaller than 2 \m, and 85 percent less than 4 ym in
geometric diameter with little difference in the spectra at the three sites.
Lundgren (1970) measured lead size-distributions in an urban area with a
Lundgren impactor and found an MMED of 0.5 pm with 5 percent of the mass in
particles greater than 17 urn in equivalent diameter.
Some Investigators have examined zinc and cadmium as well as lead dis-
tributions. Harrison et al. (1971) used a modified Andersen impactor to
sample lead and cadmium in an urban atmosphere. They found 60 percent of the
lead mass on the afterfliter (<0.2 \m in equivalent diameter) in a unimodal
distribution, while cadmium was characterized by a bimodal distribution with
very little material on the afterfilter. Lee et al. (1968, 1971) conducted
several studies measuring distributions of lead, zinc, and cadmium in
urban areas. Their results showed 68 percent of the lead, 40 percent of the
zinc, and 28 percent of the cadmium in particles smaller than 1 pm In
equivalent diameter. Lee and von Lehmden (1973) also summarized the MMED
from several researchers and found that lead has consistently smaller
MMED values than either zinc or cadmium. In particular, the lead MMED
range from 0.2 to 1.43 pm, zinc from 0.58 to 1.79 urn, and cadmium from
1.54 to 3.1 \m. In more recent studies, however, Lee and Goranson (1975)
have measured an Increasing trend in total particulate mass MMED for the
period 1970-1972.
26
To summarize the results of the above survey, it is apparent that air-
borne lead is mainly found in submicron particles with about one-third of the
mass being supermicron. However, zinc and cadmium have greater fractions
in the supermicron range.
ATMOSPHERIC SIZE-DISTRIBUTIONS: ISOKINETIC EXPERIMENTS
Isokinetic impactor runs were conducted using a modified Andersen
model 20-000 sampler. The modifications Involved construction of a new
orifice and two top stages for better fractlonatlon of large particles.
Stickey Teflon Temp-R-Tape (Connecticut Hard Rubber Company) was used as
an impactor substrate. It was found that the adhesive and the particles
completely dissolved in a hot solution of 1 part concentrated hydrofluoric
acid and 20 parts concentrated nitric acid. After evaporation and re-
dissolving, the resulting solution was analyzed for lead and zinc with
flame atomic absorption spectroscopy, and for cadmium with flameless
atomic absorption (carbon rod atomization).
Measurements were made on the roof of W. M. Keck Laboratories at Caltech
which is more than 300 meters from the nearest heavily traveled street, but
less than 1 kilometer from several such streets. The impactor was
positioned horizontally facing Into the wind, 150 cm above the roof
surface. At this height, resuspended roof dust was not a problem. The
wind direction was checked frequently during th£ runs, and if the direction
was found to have changed by more than 20°, the Impactor direction was
also changed. It was found that the wind direction changed little during
the experiments, although short gusts of wind varying + 45° occurred
occasionally. A flow rate of 13.2 liters per minute was used because at
this flow the orifice airspeed was approximately equal to the average
windspeed.
The first isokinetlc experiment took place during May 13-15 and
May 17-18, 1975. Sampling was from 8 a.m. to 8 p.m. dally. For this
run, the standard eight-stage Andersen Impactor was used with a modified
27
-------
orifice, but without modified stages. A Millipore HAWP 04700, 0.45-uo
pore-size filter was used as the after filter.
The second experiment occurred from July 16-20, 1975, with the same
daily sampling times as above. Two specially constructed stages were
substituted for the regular Andersen stage 0 to fractionate larger
particles. These new stages were similar in construction to the regular
units. The uppermost stage (stage "A") had 222 jets of 1/8-inch (3.17-
millimeter) diameter, while stage "B" had 110 jets of the same diameter.
Approximate values of median aerodynamic equivalent diameter (dp,-)
for stages 3 through 6 were determined experimentally with monodisperse
polystyrene latex and polyvinyltoluene particles (Dow Chemical Company).
The results at 13.2 liters per minute agreed with the specifications when
adjusted for the flow rate difference.
Based on effective mass mean diameters, a comparison between
isokinetic runs and previous nonisokinetic runs at the same site is
shown in Table 2. The data show that the fractions of particles greater
than 10 pm aerodynamic diameter collected by the impactor are greater
during the isokinetic runs. This shows the importance of isokinetic •
sampling.
TABLE 2. PERCENTAGE OF MASS GREATER THAN 10 Vrn AERODYNAMIC
DIAMETER FOR NONISOKINETIC AND ISOKINETIC RUNS
November 1972*
February 1974*
May 1975**
July 1975**
Pb(%)
5
11
17
17
Zn(%)
11
15
38
32
Cd(%)
—
28
37
nonisokinetic
-------
1.0
.8
. Zn
JULY 1975
.01
i
10
100
dp- /tin
Figure 5. Normalized size-distributions for lead, zinc, and cadmium
based on aerodynamic equivalent diameters.
Since the impactor was positioned horizontally, large particles
settled to the walls of the instrument. The losses were calculated to be
4 percent to 11 percent of the material on the top stages; consequently,
the measured upper stage masses were increased by these loss fractions. It
should be noted that values of dp5Q were not used for the stage cutoff
diameters. Instead, the calibration results of Flesch et al. (1967) were
used for stages 2 through 7. Stages A, B, and 1 were calibrated by running
the itnpactor isokinetically at the same site as the trace-metal experiments,
a few weeks later. Atmospheric particles were sized geometrically under
an optical microscope. Basing determinations on photographs by McCrone
and Delly (1973), many of the particles were observed to be spores, insect
parts, and other organic material. A density of 1 g/cm was assumed
based on experiments by Lundgren and Paulus (1974) who measured a density of
1.1 g/cm for similar particles.
Once the total mass-distributions on stages A, B, and 1 were obtained,
it was assumed that the mass of trace metals collected by each of these
stages had the same aerodynamic equivalent distribution as the total
atmospheric mass on each stage. The distributions on stages A, B, and 1
were then summed and added to the distribution from stages 2 through 7 and
the afterfilter.
A lower cutoff of 0.05 urn equivalent diameter was assumed based on
the work of Hidy (1973). There were very few particles with geometric
diameters greater than 40 urn, statistically too few to be significant, so
the upper cutoff was determined to be 40 um.
The error bars in Figure 5 represent one standard deviation from
the mean based on analytical and measurement errors, uncertainties in the
minimum and maximum diameters, and selected sampling errors. Although
uncertainties in the diameters are significant when the dp is less than
4 . \m, only "vertical" error bars are shown.
Figure 5 shows that lead is weakly bimodal and more uniformly dis-
tributed than either zinc or cadmium. The latter metals have strong
30
31
-------
modes above 10 urn; these findings are in agreement with the tire dust
studies. Large peaks at about 1 urn for all three metals are somewhat mis-
leading. Because of the nonideal shapes of the impactor efficiency curves,
the actual peak may occur at a smaller diameter; the reasons for this are
discussed at the end of this Section.
DEPOSITION OF LEAD, ZINC, AND CADMIUM
Several measurements of deposition velocities of lead in urban and
rural areas have been made. In addition, a number of investigators have
published values of flux J and airborne concentration C from which it is
possible to calculate the deposition velocity (v ) (equation 9).
Chamberlain (1974) has summarized the results of several investigations.
These and other data are presented in Table 3. All values are for dry
deposition, except the data of Cholak et al. (1968) and Chow (1970b).
Note that the data of Servant are for an inverted surface, and do not
include the sedimentation component (Servant, 1975).
TABLE 3. DEPOSITION VELOCITIES OF LEAD
Location
Wilderness area,
Northern California
San Nicolas Island
Residential
Toulouse, France
Lake Windermere,
Rural England
Other rural sites in
Great Britain
15 meters from a
freeway, Palo Alto,
California
Deposition Surface
Teflon plate
Filter paper
Deposit gauge,
Inverted
Filter paper
Filter paper
Polyethylene pan,
6 in. deep
vg (cm/sec)
0.07
0.093
0.13
<0.18
0.157-0.625
'0. 185 .
Investigator
Patterson and
Elias (1975)
Peirson (1973)
Servant (1974)
Pierson et al.
(1973)
Pierson (1973)
Atkins (1969)
(Continued)
32
TABLE 3. DEPOSITION VELOCITIES OF LEAD (Continued)
Location
10 meters from a
freeway, Cincinnati,
Ohio
400 meters from a
freeway in a park
removed from traffic,
Cincinnati, Ohio
3400 meters from a
freeway, in an "average
residential area,"
Cincinnati, Ohio
Downtown San Diego
La Jolla, suburban
San Diego
On the shoulder of
a freeway, Los
Angeles
Pasadena
Deposition Surface
Deposit gauge,
measuring both wet
and dry deposition
Teflon plate
g (cm/sec)
0.475
0.243
0.370
0.498
0.191
1.80
0.29
Investigator
Cholak et al.
(1968)
Chow (1970b)
Huntzicker and
Davidson (1975)
present study
The data of Table 3 are best understood by considering a lead particle
source emitting a known size-distribution. As this lead is transported
away from the source, depositional loss to the ground as well as other
processes will affect the shape of the distribution. Large lead particles
with high deposition velocities will be lost first. Consequently, a
smaller deposition velocity would be expected far from the source. (It is
assumed that particle growth is unimportant in the size range of interest).
This trend may be observed in Table 3. The highest deposition velocity, 1.8
centimeters per second (cm/sec) has been measured closest to the source, and
the smallest value, 0.07 cm/sec has been measured in a remote area. The
value of 1.8 cm/sec may be an upper limit due to nonisokinetic sampling.
33
-------
The value of 0.29 cm/sec for Pasadena was obtained by dividing the
measured lead flux onto a flat Teflon plate by the total airborne lead
concentration measured by the Impactor and afterfilter. It should be
possible to calculate this flux, and also zinc and cadmium fluxes, using the
size-distributions of Figure 5 and proper deposition velocities for each
size range.
Many researchers (Chamberlain, 1966; Islitzer and Dumbauld, 1963;
Sehmel, 1972; Sehmel et al., 1973) have shown that deposition velocities
over rough surfaces are generally greater than sedimentation velocities
for similarly sized particles. This effect is especially evident at high
wlndspeeds. Impaction, turbulent deposition, and eddy diffuslonal
deposition on the roughness elements are responsible for the high deposition
rate when compared to sedimentation.
For a relatively smooth surface, however, Sehmel (1973) has shown
that sedimentation may be controlling. In particular, v is shown to be
equal to sedimentation velocity (v ) for a windspeed of 2.2 meters per
second (in/sec) for particles larger than 0.3 ym in equivalent diameter. A
smooth brass surface was used for his study.
Three experimental surfaces were set up on the roof of Keck Laboratories
to determine if slight changes in surface roughness affected lead deposition.
It was found that lead fluxes to a flat Teflon plate, a fine sand surface
(M5.38 mo grains), and a coarse sand surface (M). 64 mm grains) differed by
less than 15 percent. On the basis of these results and Sehmel'a findings,
it was decided that sedimentation velocities can be used in an estimated
calculation of trace-metal deposition on a flat Teflon plate.
The deposition flux can be calculated from mass concentration
distributions for particles of a given size. The fluxes calculated from
data presented in Figure 5 are listed in Table 4. The fluxes may be
compared with the measured deposition on the flat Teflon plates exposed
simultaneously with the July 1975 impactor run. The experimental fluxes
given represent average values for sticky and nonsticky Teflon surfaces.
34
Fluxes calculated from the May 1975 impactor experiment are also listed.
Care must be taken when comparing the calculated May 1975 fluxes with
the July 1975 measured values. As shown in Table 5, the airborne
concentrations of the trace metals were different during the two
experiments. Note that the measured lead flux, 31.7 nanograms per
square centimeter per day (ng/cm /day), is greater than the lead flux
measurement given in Figure 4 of only 22 ng/cm /day. The difference
exists because the above experiments measured daytime fluxes. The
experiments corresponding to Figure 4 measured overall fluxes, during
both day and night.
TABLE 4. CALCULATED AND MEASURED FLUXES TO FIAT SURFACES
Measured Fluxes
Jtop
Pb 31.7
Zn 12.0
Cd 0.238
for the July
(ng/cm /day)
bottom
7.0
3.82
0.085
1975 Experiment
Jsed
24.7 + 11%
8.18 + 12%
0. 153 + 58%
Calculated
(ng/cm
Jsed
July 1975
29.2 + 9%
6.5 + 11%
0.224 + 25%
Fluxes
/day)
Jsed
May 1975
28.9 + 6%
7.7 + 10%
0.448 ± 10%
TABLE 5.
AIRBORNE CONCENTRATIONS DURING THE TWO
ISOKINETIC IMPACTOR EXPERIMENTS
Pb
Zn
Cd
July 1975 (ns/m3)
1254 + 3%
127 + 6%
4.1 + 15%
May 1975 (ng/m3)
1160 + 2%
131 + 9%
11.6 + 5%
Table 4 shows measured depositions on both upward-facing (top) and
downward-facing (bottom) plates. Both sticky and unstleky surfaces were
used; no significant differences in deposition were observed.
35
-------
The bottom plate-fluxes are considerably smaller than the top plate
fluxes. This agrees with the results of Gregory (1961),'who found a
greater deposition of 32-vnn diameter lycopodlum spores on the top of a
glass mlcrosllde than on the bottom, for small wlndspeeds.
In a separate experiment at the same site, the number of large
particles with geometric diameters greater than 10 pm which had deposited
on these flat plates was determined using an optical microscope. The
ratio of large particles on the bottom plate to large particles on the top
plate was 0.044. Thus It Is likely that the trace metal deposition on
the bottom plate was primarily due to small particles.
There are data in the literature which compare small particle
deposition on upward- and downward-facing surfaces. Bullas (1953) has
measured the deposition of fission-product activity from nuclear test
explosions on the top and bottom of a flat surface. This activity is
generally attached to condensation nuclei with diameters of less than
0.1 urn (Chamberlain, 1960). He measured a value of 0.7 for the ratio of
bottom surface deposition to top surface deposition. Megaw and Chadwick
(1957) produced a uranium fume and measured the deposition on top and
bottom flat surfaces. The fume consisted of 83 percent by mass small
particles with diameters of less than about 0.2 |0m, but with some supermicron
material. Their value of the ratio was 0.6.
These values are somewhat less than unity, which is consistent with the
data and observations of Sehmel (1973). He found that gravity assists the
turbulent transport of small particles on horizontal upward-facing surfaces by
forcing the particles downward into regions of smaller eddies. This results
in slightly greater fluxes on top surfaces than on bottom surfaces.
As an approximation, it is assumed in this analysis that the deposition
of small trace-metal particles is equal on top and bottom surfaces. Thus
the net sedimentation flux, sed, is equal to the top plate flux minus the
bottom plate flux. These values are also shown in Table 4.
36
It may be argued that brief gusts of wind in excess of the average
1.4 m/sec may be responsible for much of the deposition. This would
especially apply to the large mass of submicron particles, since small
particles have deposition velocities greater than sedimentation velocities
in moderate winds. However, the maximum wlndspeed recorded during the
experiment was only 4.4 m/sec. When the results of Sehmel (1973) for a
6.7 m/sec windspeed are used to predict deposition velocities of the sub-
micron particles, it is found that the flux of these particles is indeed
greater than the sedimentation flux. But either value is negligible
compared to the supermicron flux.
It may also be argued that turbulent deposition of large particles
causes increases over sedimentation deposition, even for smooth surfaces.
Sehmel (1973) has shown that this effect Is not noticeable for a 2.2 m/sec
windspeed, although it becomes important at 6.7 m/sec. Because of the low
windspeeds measured during the experiment, turbulent deposition is
believed to have been unimportant.
EXPERIMENTAL ERRORS
All uncertainties in this study represent one standard deviation about
the mean. Each measurement x. is assigned an uncertainty a , and the
Xi
uncertainties are compounded according to:
li
(13)
where f is any function calculated from the various x. 's, and af is the
standard deviation of f.
Measurements of the blanks are subtracted from all atomic absorption
readings, and uncertainties of 100 percent are assigned to the blank values.
In addition to the usual errors associated with atomic absorption, the
error bars of Figure 5 also include the effects of nonideal impactor
efficiency curves and errors in counting and measuring particles under the
37
-------
optical Blcroccope. The uncertainties In the calculated fluxes are also
determined froa the above errors.
Other errors also affect the size-distribution measurements and the
flux calculations. Some of these Include wall losses and bounceoff in
the impactor, nonlsokinetlc sampling, and using the Figure 5 stairstep
distribution to approximate the actual distribution. These other errors
are estimated below, but are not included in the error bars and uncertainties
In tables and figures.
Wall losses in the Andersen impactor have been discussed by May (1964)
and by Rao (1975). May found that the most significant wall losses occurred
on the sieve above the top stage. When the impactor was calibrated in this
study with atmospheric particles, the size-distribution of particles
depositing at the center of the top stage sieve was measured. It was found
that a significant mass deposited that was mostly soil-dust particles
with geometric diameters greater than 40 urn. It is estimated that about
16 percent of the mass collected by stages A, B, and 1 deposited in this
area. However, it is not likely that this extra mass included many trace-
metal-containing particles because large soil-dust particles are poor
scavengers.
May (1964) also predicted that losses result as large particles attempt
to pass around one stage and proceed to lover stages. The size of this error
has not been evaluated.
Rao (1975) found significant wall losses for large particles, especially
near jet entraces on the top stages. He also found bounceoff to be a
problem. As expected, the most significant bounceoff occurred when solid
particles Impacted on stainless steel plates. Sticky surfaces collecting
liquid particles had negligible bounceoff. Similar results have been
reported by Hidy (197.4). To minimize bounceoff in the trace-metal
experiments, sticky Teflon was used. However, the actual errors are
unknown.
38
The errors due to nonlsoklnetic sampling can be estimated for
windspeed and wind direction data obtained during the impactor runs. For
the July 1975 experiment, the average windspeed vas 1.4 m/sec while the
inlet airspeed was 1.8 m/sec. According to Watson (1954), the ratios
of measured-to-actual airborne concentrations for these conditions are 96
percent for 12-ym, 90 percent for 31-imi, and 84 percent for 37-vttn
equivalent diameter particles. Since the wind was gusty with speeds
different from the average 1.4 m/sec, the true errors may be greater.
Errors due to the impactor not always facing into the wind can also be
estimated. Frequent checks on the wind direction were made, and whenever
the wind shifted by more than 20 , the impactor was realigned. A record
of these realignments shows that the impactor could have been misaligned
by about 30 for as much as one-fourth of the time during the run. For a
windspeed equal to the inlet airspeed, Watson has predicted the ratios of
measured to actual concentrations for a 30 misalignment. His values are
92 percent for 12-ym and 85 percent for 37-ym equivalent diameter
particles. The sampling errors for the May 1975 run are slightly smaller
than for the July experiment.
Calculations of the flux J have been based on the approximate stairstep
distribution of Figure 5. By dividing each size range into several smaller
stairsteps, so that the size-distribution is closer to a smooth curve, an
estimate of the error is obtained. This was done using the raw data from
the particle-size measurements which were accurate to + 0.4 ym for a dp of
less than 4 pm. The error was calculated to be less than 1 percent.
For the particles on the lower impactor stages which were too small
to be counted, the maximum error in calculating the flux was estimated.
It was assumed that all of the mass was at one of the dp endpoints of a
stage rather than uniformly distributed across the dp range. The
error was calculated to be 3 percent of the total flux.
Another possible source of error is that the particles may change
size as they deposit on the impactor plates. The assumption that the
39
-------
size-distributions of trace metals based on aerodynamic equivalent
diameters are the same as the total distributions measured with the optical
microscope may also be a source of error.
ANDERSEN IMPACTOR CUTOFF DIAMETERS
Generally it Is assumed that the particle sizes on an impactor stage
vary from' one dp,, value to the dp,., value of the next stage. This is
strictly true only for an ideal impactor where there Is no overlap of
particle sizes from one stage to the next. However, real impactors such
as the Andersen sampler are highly uonideal. The calibration curves of
Flesch et al. (1967) were used to estimate the cutoff diameters for each
stage, and also the uncertainty in the cutoffs due to the nonideal nature
of the curves.
The size-distributions of Figure 5 assume a uniform mass distribution
with respect to log dp on each impactor stage. This yields only a stairstep
approximation to the true distribution, but there is no physical basis for
interpolating values between the cutoffs. To be consistent with the uniform
mass assumption, it is assumed that a uniform mass distribution impinges
upon an impactor stage with a known efficiency curve such as the one .shown
in Figure 6.
All particles greater than dp will be captured by the stage, while
all particles smaller than dp . will pass the stage. Particles between
mln
dp and dp will have various collection efficiencies depending on the
mln max
shape of the curve.
The curves of Flesch et al. (1967) have been used to determine the
dp cutoffs and odp . Values are shown in Table 6. For the size-dis-
m m
tributlons of Figure 5, the dp and adp values were recalculated for a
flow rate of 13.2 liters per minute. The curves are from Flesch et al.
(1967) for the calibration flow rate of 28.3 liters per minute. The
values for stage 7 (no.t studied by Flesch et al.) were obtained by
40
1.0
u
UJ
u
0.5
dPrain. dP50
PARTICLE DIAMETER (dp)
Figure 6. Impactor stage collection efficiency as a
function of particle diameter dp, for a
typical stage.
assuming the same shape efficiency curve as stage 6 but with a stage 7 dp,.
value according to Andersen specifications.
TABLE 6. VALUES OF MASS MEAN CUTOFF DIAMETERS dp
AND THE STANDARD DEVIATIONS FOR THE °
ANDERSEN IMPACTOR EFFICIENCY CURVES
Impactor
Stage
2
3
4
5
6
7
dp.C^
6.63
3.78
1.98
1.06
0.80
0.53
odpm(^)
2.25
1.51
0.74
0.50
0.43
0.29
41
-------
Inspecting the standard deviations, it is clear that the efficiency
curves are highly nonideal. The large particle stages are in fact more
nonideal than the lower stages. This is because the Reynolds numbers in
the jets decrease in the upper stages, hence the air velocity profiles
across the jets are more nonuniform (Marple and Liu, 1974; Marple et al.,
1974). Thus a particle being carried through the center of a jet has a
different velocity than a particle near the jet wall and hence a different
impactlon efficiency. Stages A, B, and 1 have the lowest Reynolds numbers
and the greatest standard deviations which necessitates calibration with
atmospheric particles.
The standard deviations for stages 6 and 7 are greater than 50 percent
of the dp values (see Table 6). If there is a submlcron peak between
0.l-\m and 0.5-vm equivalent diameter measured by Hering (1975), then a
sizable fraction of these small particles may deposit on stages 6 and 7.
Thus the large peaks at 1 \m in Figure 5 may be influenced by submicron
material.
DISCUSSION OF TRACE-METALS DEPOSITION STUDIES
The airborne size-distributions of lead, zinc, and cadmium at Caltech
in Pasadena, California, are bimodal. The upper mode which Is located
above 10 (jm in equivalent diameter, is considerably stronger for zinc and
cadmium than for lead.
Comparing the calculated and measured sedimentation fluxes shows that
reasonable agreement has been obtained. Because the sedimentation flux
Is about 70 percent of the total top plate flux, it is concluded that
sedimentation is the dominating deposition mechanism for these trace metals
at. this site
The size-distribution can be used to predict the mass deposition of .
these trace metals on a flat surface under ambient conditons. Using
sedimentation as the mechanism for deposition, it is calculated that over
95 percent of the deposition mass is due to the fraction of particules
42
greater than 10 pm in aerodynamic diameter. The deposition mass from
sedimentation is about 70 percent of the deposition mass from all
mechanisms on an upward-facing flat plate for each of the trace metals.
The percentage of particles greater than 10 um in aerodynamic diameter
is 17, 32, and 37 percent for lead, zinc, and cadmium, respectively. On
the basis of the deposition data and the size-distribution measurements,
it is concluded that small particles play only a minor role in the
deposition process. Rather, sedimentation of a few large particles
dominates the mass deposition at this site for these trace metals.
The deposition velocity of a species may decrease as the species
moves away from its source, if the depositional loss of particles from the
atmosphere is faster than particle growth. The deposition velocity of lead,
for example, decreases from 1.8 cm/sec near a freeway to 0.07 cm/sec in a
remote area.
Although sedimentation of large particles may dominate mass deposition
near the source, this may not be the case far from sources. Once the large
particles have deposited out, the smaller material may remain airborne for
considerable periods of time. These particles may eventually deposit by
diffusional deposition.
In some cases, attachment of small particles to larger particles may be
important. For example, the lead size-distribution near long stretches of
rural highways may be dominated by the very small particles produced by
steadily cruising automobiles. Far from the roadway, attachment to large
dust particles may Increase the lead MMED.
The above analysis applies only for deposition on smooth surfaces.
Deposition velocities on rough natural surfaces are likely to be greater
than sedimentation velocities, even for large particles.
43
-------
SECTION IV
EXAMPLES OF MASS BALANCES
AUTOMOTIVE LEAD IN THE LOS ANGELES BASIN
A material balance carried out on automobile-emitted lead in the Los
Angeles Basin, was based on estimates made of the daily consumption of
lead by automobiles and the amounts exhausted to the atmosphere, deposited
on the land and roadways, and advected out of the Basin. The mass flows
are based on new measurements of atmospheric lead concentrations,
particle size-distributions, the surface deposition of lead, lead in
surface water runoff, and data from the literature. The area for which
the calculation is made includes the major population centers of Los
Angeles and Orange Counties.
For a conservative species, such as lead, the terms appearing in the
material balance are well defined with respect to a given geographical area.
As examples, the total lead emitted to the atmosphere, lead deposition
on surfaces, and the quantity advected by the winds past the borders of
the region are all exact quantities for any given time period. Such
quantities can be estimated as shown below, but their values are only
approximate at this time. The estimates will, however, be useful in a
number of possible applications. The relative contributions of Los Angeles
and nearby sources to atmospheric lead concentrations in regions downwind
of the Basin can be estimated. Similar calculations can be made on the
lead input to coastal waters from various sources. Such results should be
of value to making policy decisions related to nondegradation of air
44
quality and to water quality standards. It will also be possible to
estimate the flow of certain other trace pollutants from the lead balance,
Consumption of Lead
Although detailed information on the consumption of alkyl lead
additives is not available, the lead consumption rate for Los Angeles can
be estimated using known data. In 1972, the average distribution rate of
taxable gasoline for the State of California was 104 million liters per
day (California State Board of Equilization, 1972). During the same year
Los Angeles and Orange Counties accounted for 41.4 percent of the auto-
mobiles, motorcycles, and trucks in California (Carey, 1973). If it is
assumed that gasoline consumption is proportional to the number of motor
vehicles, the consumption rate for the Los Angeles region is 42.9 million
liters per day. The average concentration of lead in Southern California
gasoline for the winter 1971-72 was 0.56 + 0.06 grams per liter (g/1).
This gives an average 1972 lead consumption rate of 23.7 + 2.4 metric
tons per day (tons/day). (For bookkeeping purposes, one more significant
figure than justified is carried).
Estimates of the average concentration of lead in Southern California
gasoline were derived from 1971-1972 U.S. Bureau of Mines data (Shelton,
1972). The average premium gasoline contained 0.73 g/1 of lead, the
average intermediate grade gasoline 0.56 g/1, and the average regular
grade 0.40 g/1. A composite value for so-called "third-grade" gasolines
was subdivided for this analysis into "low lead" gasoline with 0.12 g/1
and "unleaded" with 0.01 g/1. The usage of this subdivision may not
necessarily correspond to what one buys at a gasoline station. For
example, many actual "low lead" gasolines are probably classified by
the Bureau of Mines as "regular."
According to the Los Angeles Air Pollution Control District, 62.9
percent of the gasoline sold in Los Angeles County in the summer of 1972
was premium. An increasing fraction of low compression vehicles on the
45
-------
road and the relative increase In the consumption of lower lead gasoline
limits the validity of this percentage to 1972. The remainder is divided
among Intermediate, regular, "low lead," and "unleaded" grades. Because
the.intermediate grade lead concentration falls within the range of
concentrations found in regular gasoline, and because of the low volume
sale of "unleaded" gasoline in 1972, little uncertainty is introduced by
not including these grades in this study. Thus, the two cages considered
represent upper and lower limits for lead consumption; 62.9 percent
premium and 37.1 percent regular, and 62.9 percent premium and 37.1
percent "low lead." The average lead content from these two cases is
0.56 + 0.06 g/1 where the uncertainty refers to the two possible extremes.
The Los Angeles County Air Pollution Control District (Lemke, 1971)
estimates that during 1972 350 tons/day or about 5 x 10 liters/day of
gasoline was lost by evaporation from automobile fuel systems and during
gasoline handling operations. If the tetraethyl lead fraction does not
change during evaporation, then approximately 0.3 tons/day of lead in
tetraethyl lead vapor was emitted into the atmosphere, and the remainder,
23.4 tons/day was consumed by automobiles.
Nature Of Lead Emissions
Hirschler et al. (1957, 1964) showed that the size and amount of
lead-containing particles were sensitive functions of driving mode. At
cruising speeds the exhaust fraction varied between 14 and 54 percent.
During full throttle acceleration, however, large amounts of lead were
reentrained from the exhaust system giving exhaust fractions up to 200
percent of the input. Subsequent studies by Mueller et al. (1964),
Ter Haar et al. (1972), and Hablbi (1970, 1973) confirmed Hirschler's
general results. Ter Haar et al. (1972) found that the exhaust fraction
was small for automobiles with new exhaust systems but increased with
age; this indicated that a break-in period of several thousand miles
was necessary before.the exhaust system deposits stabilized. Thus an .
accurate picture of typical lead emissions under consumer conditions
can only be obtained by monitoring lead emissions for many thousands of miles.
Hirschler et al. (1957, 1964) studied three automobiles over periods
which included both uncontrolled suburban driving and programmed chassis
dynamometer tests. They found that three cars with 27,000, 19,300, and
9,800 accumulated miles (43,000, 31,000, and 16,000 km) retained 21.2
percent, 27.5 percent, and 23.1 percent, respectively, of the input lead.
Ter Haar et al. (1972) conducted a similar study on one car, but also
attempted to construct a complete mass balance for the lead by collecting
the exhausted lead in a cyclone separator and total filter. After 12,000
miles (19,300 km) over a fixed route, 30 percent of the input lead
remained In the oil, oil filter, and exhaust system, and 54 percent had
been exhausted. Ter Haar et al. (1972) speculated that the remaining 16
percent was lost during handling of the filter and exhaust system.
For an automobile operated under simulated consumer test conditions,
Habibi (1973) found an emission rate of 89 percent between 20,000 and
33,000 accumulated miles (32,000 and 53,000 km). However, Habibi's
results for these mileages are probably not representative of the total
vehicle population since they do not take Into account the break-in periods
for new cars and cars with new mufflers.
To estimate the amount of lead exhausted, the retention factors
(I.e., percent of lead remaining in the car) of Hirschler et al. (1957,
1964) and Ter Haar et al. (1972) are assumed representative of the total
vehicle population. A retention factor of 30 percent Is used for the Ter
Haar results without trying to account for the missing 16 percent. This Is
a consistent application of the Hirschler and Ter Haar data since
Hirschler et al. measured only the lead remaining in the car and did not
construct a total mass balance. The average retention factor for the four
automobiles sampled was 25+4 percent. Thus, of the 23.4 +2.4 tons/day
of lead burned, 5.8 + 1.1 tons/day is retained In the cars, and 17.6 +
2.6 tons/day (by difference) Is exhausted to the atmosphere. (Because of
the changing nature of gasoline consumption, this exhaust rate applies
strictly only to 1972.) The use of the Hirschler and Ter Haar retention
factors implies muffler and exhaust system changes at about 27,000-mile
47
-------
(43, SCO-ton) intervals although the actual interval may be longer. Such a
change will restore the automobile to a low emission state and will entail
another break-in period. Firm data are not available to evaluate these
effects, however.
Brief (1962) has shown that auto exhaust lead contains both a particulate
fraction and an organic, vapor-phase function. Measurements on eight pre-
1961 European and English cars showed that the vapor-phase component was
about 12 percent of the particulate component. Because of his sampling
scheme, however, the particulate lead measured by Brief probably consisted
primarily of particles smaller than about 9 \m which according to Habibi
(1973) (see below), make up 43 percent by weight of the particulate exhaust.
Thus, if Brief's results can be applied to Los Angeles automobiles,
approximately 0.9 tons/day of organic, vapor-phase lead and 16.7 tons/day
of particulate lead are exhausted. Recent measurements of particulate and
vapor-phase organic lead near roadways by Skogerboe (1974)* also indicate
a significant organic lead contribution. The fate of the vapor-phase
lead is discussed later. Lead emissions from nonautomotive industrial
sources are small C\<0.3 tons/day) and are not included in the mass balance.
Of the lead which remains in the car, one-third of one-half is in
the oil and oil filter (Hirschler et al., 1957, 1964; Ter Haar et ai:, 1972.)
The fate of used motor oil is not well documented. There Is some disposed
directly into the sanitary sewer system, uncontrolled disposal onto the
ground, and limited reclamation for reuse (Bowen, 1972.)
Lead Measurements
The material balance method developed in this Section is based on
measurements of atmospheric particle size—distributions, surface deposition
fluxes and atmospheric lead and carbon monoxide concentrations. Measurements
Skogerboe, R. K., private communication (1974).
48
were made to provide data which were relative to the reference year 1972,
and based on a consistent set of experimental and analytical techniques.
Earlier measurements of particle size-distributions (Robinson and Ludwig,
1967), airborne lead concentrations (Tepper, 1971; National Air Surveillance
Networks, 1973; Public Health Service, 1965; Colucci et al., 1969), lead
deposition on Avena sativa (wild oats), (Rabinowitz, 1972), and carbon
monoxide concentrations (Public Health Service, 1965; Lemke, 1971;
Colucci et al., 1969) have been reported for the Los Angeles area.
Lead deposition was measured at Pasadena for ten different 1- or 2-
week sampling periods between November 1972 and February 1974. A 1- week
synoptic measurement of deposition was made at Pasadena and five other
locations in the Los Angeles Basin during August 1973. Deposition measure-
ments were also made at three locations near a freeway in May 1973 at two
coastal islands, and at four sites each along the coast, in the Mojave
Desert, in the San Gabriel-San Bernardino Mountains, and in the Coachella
Valley during the summer of 1973. During each of these measurement periods,
lead deposition in Pasadena was also measured. Size-distribution measure-
ments were made at Pasadena and a freeway. Size-distribution and
atmospheric lead concentrations were also measured at five sites in the
Los Angeles Basin during the 1972 and 1973 phases of the California
Aerosol Characterization Experiment (ACHEX) (Hidy, 1973.)
Experimental errors are expressed as one standard deviation about
the mean and are related only to variations about the mean unless noted
otherwise. These errors are not related to individual uncertainties in
each separate measurement. Uncertainties for derived quantities are
compounded in the usual manner.
Removal Of Lead From The Atmosphere: Deposition
Of the 16.7 tons per day of particulate lead which is exhausted, most
is in the form of lead halide particles; however, other chemical species
49
-------
have been identified. Depending on the particle size, a number of environ-
mental pathways are available. For example, very large particles (dp > 10 vm
settle rapidly. (Unless otherwise specified, all particle diameters are for
the aerodynamically equivalent, unit-density spheres.) Smaller particles
may remain airborne for a longer period, but some eventually deposit in
the urban area by convective diffusion to surfaces. The mechanisms for
particle removal have been discussed by Chamberlain (1967a, 1967b) and
Sehmel (1972) for relatively well-defined smooth and rough surfaces. A
knowledge of the particle size-distribution is important in evaluating these
effects.
The size distribution of auto exhaust lead aerosol has been studied
by many investigators (Hirschler et al., 1957; Hirschler and Gilbert, 1964;
Mueller et al., 1964; Ter Haar et al. , 1972; Habibi 1970, 1973; Moran and
Manary, 1970; Lee et al., 1971; Ninomiya et al., 1971; Ganley and Springer,
1974). Only Habibi (1973) and Ter Haar et al. (1972) attempted to simulate
actual driving conditions, and of these, the study by Habibi vas most
detailed. Consequently, much of the following discussion is based on
Habibi's results.
Habibi (1970) sampled auto exhaust aerosol at the end of a 12-meter
long wind tunnel with an isokinetically operated cascade impactor. Coarse
particles deposited in the wind tunnel were also measured and assigned
to the first impactor stage. For an automobile operating solely on a chassis
dynamometer programmed to the 1968 Federal Mileage Accumulation Schedule,
the mass median diameter of the particulate lead Increased from about
1 \m at 5,000 accumulated miles (8,000 km) to greater than 15 \m at
28,000 miles (45,000 km). At this latter mileage, 57 percent of the mass
of lead was associated with particles larger than 9 ym in diameter. A
car which had been driven on the road under typical consumer conditions
for 15,000 miles (24,000 km) followed by about 18,000 miles (29,000 km)
on the programmed chassis dynamometer also gave a lead aerosol with 57
percent by weight in particles larger than 9 ym in diameter and a mass
median diameter greater than 15 ym.
50
Ter Haar et al. (1972) tested 26 cars with 17.000 to 92,000 miles
(27,000 to 148,000 km) of service and concluded that 55 percent of the
emitted lead was in "coarse" particles (i.e., particle diameters greater
than about 5 urn). For an automobile operated for 12,000 (19,300 km) miles
on a mileage accumulation route, Ter Haar et al. also found that 58 percent
of the emitted lead was as course particles. (As noted above, however,
16 percent of the input lead could not be accounted for.) Ninomiya et al.
(1971) measured the size-distribution of auto exhaust aerosol and found that
approximately 20 to 30 percent by weight of the lead was a "coarse material"
(500 ym < dp < 5,000 ym) for a dynamometer cycle consisting of a cold start
followed by four Federal test procedure driving cycles.
For the mass balance analysis presented here, it is assumed that the
particle size-distribution of auto exhaust lead at 33,000 accumulated miles
(53,000 km) measured by Habibi (1973) is typical of Los Angeles cars.
From the known age distribution of California automobiles and the mileage
accumulation rate of Los Angeles automobiles (Lees et al., 1972), an
average accumulated mileage of 57,000 (92,000 km) is estimated. This
distribution is plotted in Figure 7 along with lead size-distribution
taken at a receptor site in Pasadena which was not in the Immediate
influence of traffic. Dp is the aerodynamic, unit-density particle
diameter. In the receptor site distributions, only 2 to 7 percent of the
mass is in the greater than 9-ym fraction as compared to 57 percent for
the auto exhaust distribution. This is consistent with the size-
distributions determined at receptor sites in the ACHEX study (Hidy, 1974)
which also show very little mass (<5 percent) above 9 urn. In Figure 8,
differential mass distributions for lead aerosol at a site 1 meter from a
freeway and at the Pasadena receptor site show that the large particle mode
Dp > 7 )Jm) at the freeway is severely attenuated at the receptor site.
Size-distribution measurements by Daines et al. (1970) also show a
decrease in large particle lead with increasing distance from a highway.
The dashed lines in Figure 8 are based on the assumption that the smallest
particle size is 0.01 urn and the largest is 50 ym.
51
-------
dp(|»m) .„ .
0.4 -
• % OF MASS IN PARTICLES < dp
Figure 7. Cumulative mass distributions for lead aerosol in auto
exhaust and in ambient air at Pasadena. (Dp is .the
aerodynamic, unit-density particle diameter.)
The difference between the source and receptor size-distributions is
due in part to the rapid deposition of very large particles near the
roadway. Habibi (1970, 1973) found that approximately one-half to two-
thirds of the greater than 9-ym fraction deposited within 7 meters of the
automobile exhaust pipe in the wind tunnel experiments. In this current
analysis, it is assumed that all of the greater than 9-ym fraction deposits
on or near the roadway in what is termed "near" (source) deposition.
Although a cutoff at 9 ym is somewhat arbitrary, Heichel and Hankin (1972)
found that lead-containing particles deposited on trees adjacent to a.
heavily traveled road ranged in aerodynamic diameter from 7 to 32 ym with a
mean of 17 ym. The near deposition fraction is 57 percent of the
exhausted lead for which an uncertainty of + 10 percent is assigned to
52
dp (urn)
Figure 8. Differential mass distributions for lead at
1-meter from a freeway and for Pasadena.
account for variations among automobiles and muffler changes at various
mileages. This amounts to 9.5 + 2.2 tons/day.
Deposition measurements at 1, 30, and 150 meters from a freeway, at
which the traffic flow was unidirectional and slightly downhill, gave
a factor of 15 decrease between the sites at 1 and 30 meters. The integrated
deposition over this 150-meter strip could account for only about 10 percent
of the near deposition expected from average emissions based on the above
considerations. However, Habibi (1973)* has pointed out that most of the
Habibi, K., private communication (1973).
53
-------
large particle emission probably occurs at locations such as freeway
onraMps or uphill stretches where heavy accelerations occur frequently.
Deposition data near such sites were not acquired. Other studies (Cahill
and Feeney, 1973; Dalnes et al., 1970) have shown that considerable
variation in the transport of large particles of lead away from the freeway
can occur depending on road characteristics and wind conditions.
It is convenient to designate lead which deposits at large distances
from the source as "far" deposition. The average Pasadena deposition for
different periods between November 1972 and February 1974 was 45 + 11
eg/cm /day. The deposition fluxes measured at different sites during the
synoptic urban sampling period have been normalized to the Pasadena flux
measured during the same period and are shown in Figure 9. These
normalized deposition fluxes are designated "deposition factors." The
average deposition factor for the basin was 1.0 + 0.4. This value excludes
the coastal sites which are upwind from sources in the Los Angeles Basin
most of the time. In Figure 9 the squares correspond to current surface
deposition measurements, the open circles to the wild oats (Avena eati-va)
data (Rabinowitz, 1972), and the triangles to the airborne data (Tepper,
1971); National Air Surveillance Networks, 1973; Hidy, 1973). The solid
circles indicate specific locations, e.g., Pasadena. The southern
boundary of the basin is taken to be the Pacific Coast and the northern
boundary the first crest of the San Gabriel Mountains (e.g., Mt. Wilson.)
Related data such as the atmospheric concentrations of lead and the
concentrations of lead on the tops of wild oats (Avena eativa) can be used
to check whether current deposition measurements are characteristic of the
tos Angeles Basin. Studies by Motto et al. (1970), Dedolph et al. (1970),
and Rabinowitz (1972) have shown that lead in the tops of grass comes
primarily from lead aerosol while the soil contributes only 2 to 3 parts
per million (ppm) (dry weight) of lead. Deposition on horizontal surfaces
and on oats depends primarily on the atmospheric concentrations of lead
and is relatively insensitJ-e to the particle size-distributions and wind
conditions (Chamberlain, 19t7a, 1967b; Sehmel, 1972.) Thus deposition
54
MT" D
13 SAN GABRIEL MOUNTAINS
O 1.6
00.75
PUENTE HILLS
_
a THIS WORK
O WILD MT IAKNA SATIVAI OAT*
A AIRBORNE DATA
NEWPORT REACH-**-60
•'Q
Figure 9. Distribution of deposition factors in the Los Angeles Basin.
factors based on normalized concentrations of atmospheric lead and lead on
the tops of wild oats should be equivalent to flat surface deposition
measurements used here.
The Rabinowitz (1972) measurements of lead concentrations in wild
oats between December 1971 and January 1972 at many sites in the Los Angeles
area represent the accumulation of lead over the whole growing season.
Lead is not washed off Avena by rainfall (Rains and Thornton, 1970). Tepper
55
-------
(1971) has reported airborne lead concentrations at eight sites In the
Los Angeles area for 1968-69 and the National Air Surveillance Networks
(1973) for four sites for 1969. Airborne lead concentrations for two
sites In the Los Angeles Basin during 1972 have been measured by ACHEX
(Hidy, 1973). The airborne data were normalized to measured Pasadena
values. The oats data were first normalized to measurements made in
central Los Angeles and renormalized to Pasadena on the basis of the
airborne lead data because there were no wild oats data measurements in
Pasadena. These normalized values are shown in Figure 9.
The distribution of deposition factors in Figure 9 indicate that
current deposition measurements are consistent with measurements by others.
The deposition flux can be calculated from the product of the average
2
Pasadena deposition (45 + 11 ng/cm /day), the Basin-wide deposition factor
(1.0 + 0.4), and the area of the Los Angeles Basin (4,430 square kilometers).
The resultant far deposition is 2.0+ 1.0 tons/day.
This value assumes that the Los Angeles Basin is a smooth surface.
Vegetation and manmade structures will alter this number, but the magnitude
of these effects is difficult to evaluate. The deposition on grassy
surfaces can be estimated from the deposition velocities of Sehmel et al.
(1973) for 0.7-cm grass and an 8-kilometer per hour (km/hr) wind. If the
average atmospheric concentration is taken to be 2.4 pm/m (the average for
all measurements In this study) and the Pasadena size-distribution assumed
to be typical of the Basin, then the estimated lead deposition for grass
is about 40 ng/cm (of land area) per day, which is not significantly
different from the experimental value for the Teflon surface. Deposition on
vertical surfaces will also contribute to the total far deposition. The
vertical surface of buildings is about 25 percent of the total land area,
and only about one-fourth of this is exposed to the wind at any given
time. Thus, the additional deposition on buildings is probably small.
Trees will also Increase the far deposition term, but this effect is
difficult to estimate. In general, more research is necessary to clarify
these uncertainties.
56
Airborne Lead: Removal By Wind
Airborne lead an carbon monoxide, both of which result almost
completely from automobile exhaust in the Los Angeles area, are known to be
positively correlated (Public Health Service, 1965; Colucci et al., 1969).
On the time scale of the air flow through the Los Angeles Basin, carbon
monoxide is unreactive and therefore conserved. Only on very smoggy days
with high oxidant concentrations will carbon monoxide be removed at rates
on the order of 1 percent per hour by reaction with a hydroxide (Weinstock
and Niki, 1972; Wilson, 1972). To estimate the amount of lead which is
blown out of the Basin, the Basin is considered to be a continuously
stirred flow reactor. In this approximation the input and output flows of
carbon monoxide are equal, and the lead output flow can be estimated from
the relationship.
CO
(14)
where q. is the mass flow rate of species i out of the Basin and [Pb]/[CO]
is the average value of the ratio of lead to carbon monoxide at receptor
sites in the Basin. This approximation Implies a constant proportionality
between lead and carbon monoxide throughout the atmospheric mixing layer.
Although vertical profiles of lead concentrations are not available, this
assumption is reasonable because the dominant mechanism for the mixing of
both species is eddy diffusion which will disperse both the gas and the
aerosol (of which about 70 percent by weight is in the submicron fraction)
in a similar manner.
Lead and carbon monoxide concentrations at five different sites in the
Los Angeles Basin are available for 1972, 1973, and 1974 from current studies
and from the ACHEX measurements. With the exception of one sampling period,
all the lead measurements were made by high-volume sampling with Whatman
41 filters. Simultaneously, lead concentrations were sampled at 2-hour
intervals by low-volume sampling with Gelman GA-1 membrane filters. For
the 1973 ACHEX measurements, the ratio of "high-volume" lead to "low-
volume" lead was 0.86 + 0.08 where "low-volume" lead is the lead concentration
57
-------
determined by summing and averaging the low-volume filters over the high-
voluse saapling period. The high-volume filter data Is used because of
the greater possibility of systematic error (e.g., from contamination)
Involved In the handling of the low-volume filters. However, within the
accuracy of the treatment In this analysis, no significant difference
results from the choice of one or the other set of data.
The average lead-to-carbon monoxide (Pb/CO) ratios for the sites are
given in Table 7 and the average for all sites Is 0.69 ± 0.18 vig/m lead/ppm
carbon monoxide. The Pomona site is at the eastern border of the study
region and the Rubldoux site is further east and outside the study region.
Consequently Pb/CO ratios at these sites should be representative of air
leaving Los Angeles since the major air flow is toward the east and north.
The close agreement between Pb/CO at these sites and Pb/CO Implies that the
continuously stirred flow reactor formalism is a reasonable approximation.
_TABLE 7. Pb/CO AT VARIOUS SITES IN THE LOS ANGELES BASIN
Site and Number
of Sampling Days
Pasadena (10)
Pomona (7)
West Co vlna (5)
Domlnguez Hills (2)
Rubidoux (3)
Average
Pb/CO (lJg/m3
0.46
0.74
0.79
0.81
0.65
0.69
/ppm)
+ 0.18
When the Pb/CO ratio is adjusted for nonautomotive carbon monoxide
and converted to a dlmensionless weight ratio, the ratio Is 6.2 + 1.6 x
-4
10 . On the basis of seven mode cycle data, the Los Angeles Air Pollution
Control District estimates that 7,200 tons of carbon monoxide was emitted
each day by automobiles during 1972. Orange County adds an additional 20
percent (State of California, 1972), bringing the total to 8,600 tons/day.
58
An uncertainty of ±50 percent is assigned to this emission although the
uncertainty may be even larger. The resulting rate of removal of particulate
lead by advection is then 5.3 + 3.0 tons/day.
In addition to the particulate airborne lead there is also a vapor-
phase, organic cotnponent. Purdue et al. (1973) found the organic component
to be about 10 percent of the total for six American cities. Skogerboe
(1974)* measured organic fractions ranging between 4 and 12 percent at a
receptor site in Fort Collins, Colorado. The work for this current report
was conducted over a 3-day period in June 1974 and measurements of the
organic component were found to be 6 ± 1 percent of the total Pasadena
airborne lead. If this is typical of the Los Angeles region, 0.3 ton/day
of vapor-phase lead is removed by advection. The difference between the
input of organic lead (1.2 tons/day) and the output is due to photolysis of
the organic lead vapor to produce a lead-containing aerosol (Milde and
Beatty, 1959). Eventually all of the vapor-phase lead will decompose to
an aerosol.
Lead Input To The Coastal Waters
The input routes to the coastal waters may include atmospheric
deposition, rainout-washout, rainy weather runoff, dry weather runoff,
discharge of treated sewage, and direct discharge of untreated wastes.
These are all considered here with the exception of direct discharge of
untreated wastes.
Because of the lack of data on the particle collection properties of
the ocean surface, the estimate of the atmospheric deposition on the coastal
waters is based on dry deposition measurements on the coastal islands.
As noted earlier, the lead deposition rates on the standard Teflon
collector and on a crystallization dish filled with water were the same.
It is recognized that the dynamics of the ocean surface are quite different
* Skogerboe, R. K., private communication (1974).
59
-------
from chose of the crystallization dish, and wave action may alter the
particle deposition velocity at the ocean surface.
The deposition fluxes of lead on Santa Catalina Island and San Clemente
Island during the summer of 1973 are plotted as a function of distance from
the coastal baseline in Figure 10. The open squares in Figure 10 are
deposition measurements from this work, the solid square is the deposition
measurement by Patterson and Settle (1974), and the circles are the
equivalent flat-surface deposition fluxes from the Avena sativa data
(Rabinowitz, 1972). Rabinowitz (1972) has measured the concentration of
lead in the tops of wild oats on San Nicolas, San Clemente, Santa Barbara,
and Santa Catalina Islands, and along the coast. From deposition factors
calculated from these data, equivalent flat surface deposition fluxes
have been derived and are also plotted in Figure 10. A soil contribution
of 2.ppm (dry weight) for the oats data was subtracted. The coastal
deposition value in Figure 10 is the average of the surface deposition and
oats data.
•= 15
X 10
T
T
SANTA BARBARA IS.
\
SAN CLEMENTE IS.
1
> F
SAN HI
=J
_L
O
O
0 50 100
DISTANCE FROM COASTAL BASELINE (km)
Figure 10. Deposition of lead as a function of distance from the-coast
60
Patterson and Settle (1974) have measured the surface deposition of
lead on Santa Catalina Island for the 2-week period immediately following
measurements in this current study and found a flux of 1.4 ng/cra /day as
compared with a value of 3.3 ng/cm /day reported here. They also
measured the lead-206 to lead-207 (206Pb/207Pb) ratios for the Catalina
lead and the lead depositing in Pasadena and found a significant
2
difference (1.171 and 1.193 ng/cm /day, respectively). Since sources of
lead are different if the Catalina 206Pb/207Pb ratios differed by 0.3
percent from the Los Angeles lead ratios, they concluded that the lead
depositing on Catalina during their measurement did not originate in Los
Angeles air. This is consistent with the local meteorology. The prevailing
wind pattern during the year is onshore flow during the day. At night,
weak offshore drainage winds converge with the normal onshore flow. The
western edge of the convergence zone is poorly defined, but may extend as
far offshore as the eastern edge of Santa Catalina Island (DeMarrais
et al., 1965.) Thus Santa Catalina can occasionally receive Los Angeles
air, but except under Santa Ana conditions (east winds from the mountains
and deserts), the islands which are farther offshore do not receive Los
Angeles air.
Hidy et al. (1974) measured an average atmospheric lead concentration
of 120 nanograms per cubic meter (ng/m ) on San Nicolas Island (about
120 km from the coast) during the summer and fall of 1970. This corresponds
to a surface deposition rate of about 2 ng/cm /day which is in reasonable
2
agreement with the Patterson and Settle value of 1.4 ng/cm /day for Santa
Catalina and the value reported here of 2.4 ng/cm /day for San Clemente Island.
The somewhat higher value of Santa Catalina deposition in this current study
2
(3.3 ng/cm /day) may represent a slight penetration of Los Angeles air,
although no isotopic ratio measurements were made to confirm this. Early
morning east winds on Santa Catalina were recorded on 5 days during these
measurements, but not at all during the Patterson-Settle measurement. The
deposition of cadmium on Santa Catalina was also somewhat higher than on
San Clemente Island (Davidson et al., 1974).
61
-------
The interpretation of the coastal Island data assumes, then, that the
influence of Los Angeles beyond Santa Catallna Island is negligible.
Because deposition data between the coast and Catallna are not available,
a staple linear interpolation (the dashed line in Figure 10) is used to
estimate the decrease in deposition over this region. The "background"
or non-Los Angeles lead is represented by the solid line in Figure 10
and is the average of Patterson and Settle's Catalina flux and the fluxes
for the outer islands. The dashed line in Figure 10 defines a region of
influence of Los Angeles extending to about 55 km from the shore, which
for a linear coastal distance of 80 km corresponds to an area of 4,400
km . The average (dry) deposition flux over this region is 0.35 tons/day
or 120 tons/year assuming 346 "dry" (i.e., rainfall less than 2.5 mm/day)
days (U. S. Department of Commerce, 1968-73). Further measurements are
necessary to define the origin of the "background" lead.
Chow et al. (1973) and Bruland et al. (1974) have measured lead
accumulation rates in the ocean sediment of 4.7 ng/cm /day in the San
Pedro Basin (^ 30 km for the coast), 2.5 ng/cm2/day in Santa Monica Basin
(^ 50 km from the coast), and 5.7 ng/cm /day in the Santa Barbara Basin
C\. 30 km from the coast). The Santa Monica and San Pedro Basins are directly
off Los Angeles while the Santa Barbara Basin is off Santa Barbara, about
100 km northwest of Los Angeles. Although these accumulation rates are
of the same order as the deposition fluxes shown in Figure 10, their
Interpretation is complicated by various factors: the influence of sewage
outfalls, a combination of advective ocean transport and settling, and
dissolution during settling. Thus a direct comparison between these
accumulation rates and deposition measurements is not possible, but the
general agreement tends to support our interpretation of the deposition
data.
The lead input to the coastal waters associated with rainout-washout
can be calculated from the average annual rainfall over the coastal waters
and the lead content of the rainfall. The average rainfall at three
62
coastal sites and Santa Catallna Island is 30 cm per year (U.S. Department
of Commerce, 1968-1973). In 1966-67 Lazrua et al. (1970) measured the
2
flux of rainfall lead on Santa Catalina to be 25 ng/cm /cm of rain. Thus
the lead input to the region of influence defined by the deposition measure-
ments is 30 tons per year.
Although positive identification of the origin of the rainfall lead
has not been made, meterological data strongly suggest Los Angeles as the
source. During the period of heaviest rain, November through February, the
majority of the winds are from the quadrant centered on east-east northeast
with an average speed of about 18 kra/hr (DeMarrala et al., 1965). Such
winds carry Los Angeles air well out over the ocean. In fact, the region
of influence of the winds from Los Angeles may be larger than the region
of influence determined from the deposition measurements. However, no data
are available to confirm this.
In addition to the direct input by rainout-washout, rain storms will
wash lead-containing particles into the storm sewers and ultimately Into
the coastal waters. Lead in runoff results primarily from street dirt.
Soluble lead deposition on soil is immobilized by sorption on soil particles
(Zimdahl, 1972) and is not subject to significant washoff. During the winter
of 1971-1972, the Southern California Coastal Waters Research Project (SCCWRP,
1973) measured trace-metal concentrations in storm-water runoff flowing in
the concrete-lined rivers, which act as storm sewers for the region. These
rivers carry significant amounts of water only during storms. The samples
were obtained by a depth-Integrated method at various times throughout the
storms. The chemical analysis Included both the dissolved and partlculate
fractions for the metals. There was only one major storm during this period,
however, and it is difficult to arrive at a firm number for the runoff input.
Because a firm data base is lacking for this environmental pathway, a
simple model for estimating the flux of runoff lead was developed. This
model is discussed later in this Section. .It is estimated from the model
that approximately 140 tons/year is washed off the streets into the coastal
waters during the rainy season (November-April). During the dry season, a
63
-------
small amount of water flows in the streets as a result of such domestic
activities as lava watering. Runoff during this season adds another 10
tons/year.
The final contributor to the anthropogenic lead burden of the coastal
waters is the discharge of treated sewage in which the total lead emission
rate is 230 tons/year or 0.64 tons/day (Parkhurst, 1973; City of Los
Angeles, 1973; County Sanitation Districts of Orange County, 1973). The
coastal water Inputs are summarized in Table 8. The four "atmospheric"
routes (dry deposition, ralnout-washout, and dry-and wet-season runoff)
account for more than 50 percent of the total, indicating that effective
control of lead emissions to the coastal waters can be achieved only
when both sewage and "atmospheric" routes are controlled. The runoff and
sewage routes are essentially point sources of contaminants to the
coastal waters whereas the deposition and rainout-washout routes act over
a larger area of influence. It is likely that these two types of routes
will have different ecological effects.
TABLE 8. INPUT OF ANTHROPOGENIC LOS ANGELES LEAD TO THE COASTAL WATERS
Input Route
Storm Runoff
Dry Deposition
Rainout-Washout
Dry Runoff
Municipal Sewage
Tons /Year
140
120
30
10
230
TOTAL 530
Tons /Day
0.4
0.3
0.09
0.03
0.64
1.46
Transport Of Windblown Lead Outside The Basin
To investigate the fate of the 5.6 tons/day of lead which is blown
out of the Los Angeles Basin, the lead deposition was measured at four
sites each in the San Gabriel-San Bernardino Mountains north of Los Angeles
64
and in the Mojave Desert north of the mountains. The average deposition
factors in these regions were 0.07 and 0.04, respectively, which are in
general agreement with the factors calculated from the wild oats data
(Rabinowitz, 1972) for the same regions. These factors, which are
uncorrected for local sources, represent the maximum amount of Los Angeles
lead depositing in these regions. If these deposition factors are
o
applied to a 10,000-km area corresponding to the mountain and desert part
of the quadrant centered at northeast and extending approximately 150 km
from downtown Los Angeles, a deposition flux of 0.2 tons/day is obtained.
Although the source of this lead cannot be unequivocally attributed to
Los Angeles, the general wind patterns and the sparse local traffic density
strongly suggest Los Angeles as the source.
Directly to the east of Los Angeles is a semiurban region (Riverside-
San Bernardino) comprising about 2,500 km in area. Atmospheric lead
concentrations measured at two sites in this region during the 1972-1973
ACHEX study were typical of lead concentrations in the Los Angeles urban area.
From the previous discussion of lead deposition, the maximum deposition factor
for Los Angeles lead in this region is equal to the average Los Angeles
deposition factor. Maximum is specified because local sources, as opposed
to lead blown out of Los Angeles, will also contribute to the deposition.
This gives a flux of 1.1 tons/day. When the deposition in the mountains,
deserts, and coastal waters is included, this accounts for only 1.7 out
of the 5.6 tons/day which leaves Los Angeles. The difference, about 4
tons/day, is transported beyond a radius of 150 km from the city and
exceeds the total lead emissions in the surrounding semiurban and nonurban
areas as estimated on the basis of the respective motor vehicle populations
and carbon monoxide emissions (State of California, 1972) .
Roadway Deposition Of Lead And Runoff
The accumulation of lead in street dirt, the source of runoff lead,
can be approximated by a simple differential equation:
Q-Rm
65
(15)
-------
the solution of which Is:
-RAt
(16)
where m Is the mass of lead on the street, Q Is the deposition rate minus
the amount blown off the roads by wind, R is the fraction removed per unit
of time by streetsweeplng, At is the time in weeks since the last rainstorm,
and nig is the amount of lead remaining on the streets at the end of the
last storm. Equation 16 is exact only when deposition and sweeping are
continuous and constant with respect to time, but is sufficiently accurate
for this analysis. The amount of lead washed off the streets is the product
of m in equation 16 and the washoff efficiency, £, which is a function of
the rainfall intensity.
During the first major storm (December 21-22) of the 1971-72 rain
season, 9.8 tons of lead washed off the Los Angeles River and Ballona
Creek drainage basins (SCCWRP, 1973). These drainage basins account for
39 percent of the traffic density in the Los Angeles area and an extra-
polation on the basis of traffic density to the whole Los Angeles area gives
a washoff of 25 tons (Roberts et al., 1971). Because of the weak nature of
the preceding storms, it is likely that the mass loading of lead on the
streets prior to the December 21-22 storm was close to the steady-state
value (Q/R in equation 16). Thus e Q/R is equal to 25 tons.
Annual washoff can be estimated in the following manner. The rainy
season in Los Angeles extends from November to April with 8 days of rain-
fall greater than 12 mm (U.S. Department of Commerce, 1968-1973) or an
average of about one major storm per month. Actual rainfall patterns are
complex and variable. Fox a model rain-year comprised of six major storms
at 4-week Intervals, At in equation 16 is 4 weeks for all storms except the
first. Pitt and Amy (1973) found that the removal efficiency of lead by ..
streetsweepers was 51 percent for a single pass. If the street sweeping
frequency Is once per week, R is 0.51 per week. Using these values of
R, At and e ^ and neglecting m for all storms, we obtain a washoff in
K U
the first storm of 25 tons -and in subsequent storms 22 tons per storm for
66
a total of about 140 tons per year. This estimate is independent of Q/R
and is only weakly dependent of R and o_. The critical assumptions were
the representative nature of the December 21 storm with respect to washoff
efficiency and the extrapolation to area-wide runoff on the basis of traffic
density. Considerable variations in the runoff lead-flux can occur however,
and more data are needed to better describe this environmental pathway.
The deposition rate of lead on the streets can be estimated by
setting the washoff efficiency for the December 21 storm to one. If, as
assumed above, the streets are swept once a week with a 51 percent
efficiency for lead removal, the roadway deposition rate is 1.8 tons/day
or 660 tons/year. The remainder of the near deposition, 7.7 tons/day,
deposits on the land adjoining the roadways. Because of the model dependency
and the assumptions involved, these numbers must be regarded as approximate.
There is also a small runoff contribution during the dry season. We
have measured the total lead content of dry weather runoff in the Ballona
Creek drainage basin. If the result is extrapolated on the basis of traffic
density to the whole urban area, a dally flow of 0.03 tons/day or 10 tons/year
is obtained. The total input to the coastal waters via runoff Is therefore
150 tons/year. The difference between the roadway deposition and the runoff
flux Is 510 tons/year or 1.4 tons/day. This Is the amount removed by
streetsweeping. In Los Angeles this lead is dumped in sanitary landfills and
is immobilized in the same manner ae soil lead.
Mass Balance for Automobile-Emitted Lead
The flow of automobile-emitted lead is summarized In Figure 11. In
this flow diagram the flow rates are daily averages, 1.e., the flow rates
are the yearly fluxes divided by 365.
The mass balance constructed is shown in Table 9. The agreement between
input and output routes is good. This mass balance applies to dry weather
which is the situation for more than 90 percent of the days in Los Angeles.
67
-------
INPUT
23.7±2.4
EVAPORATION
0.3
RETAINED IN CAR
5.8±l.1
(1.4)
NEAR DEPOSITION
STREET
(1.8)
(RAINY)
(DRY)
SEWAGE
LAND
(7-7)
FAR
DEPOSITION
(UNO)
RUNOFF
(0.4)
0.03
0.3
(DRY
DEPOSITION)
COASTAL WATERS
0.09
RAIN
Figure 11. The flow of automobile-emitted lead through the Los Angeles Basin
(tons/day).
68
Because each of the output terms in the balance was independently derived,
the agreement between input and output indicates that all major environmental
pathways have been considered. This conclusion must be tempered, however,
by the relatively large uncertainties in each of the terms. Improvements
in the flow estimates will require better source characterization and a
better theoretical and experimental understanding of particle removal
processes.
TABLE 9. MASS BALANCE FOR AUTOMOBILE-EMITTED LEAD
Input (tons/day)
Evaporation of
tetraethyl lead 0.3
Auto Exhaust 17.6 + 2.6
(Aerosol 16.7)
(Organic Vapor 0.9)
Output (tons/day)
Near Source
Deposition
Far Deposition
Removed by Wind
(Aerosol 5.3)
(Organic Vapor 0.3)
9.5 + 2.2
2.0 + 1.0
5.6 + 3.0
17.9 + 2.6*
17.1 + 3.9
*Inclusion of industrial emissions (0.3 tons/day) increases the total lead
emissions to 18.2 tons/day.
The three output routes listed in Table 9 are not necessarily sinks.
For example, lead which deposits on the streets can be washed off by rain
into storm sewers which empty into the ocean, or lead which is blown out
of the Los Angeles Basin serves as an input to other geographical regions.
These secondary processes and their respective removal rates are shown on
Figure 11.
Data for the rainout-washout of lead over the land area of the Los
Angeles Basin are not available. Thus the values for far deposition and
wind removal are for dry weather only. Of the 18 tons/day which is
69
-------
exhausted, approximately two-thirds of the lead Is deposited within the
Basin while the remaining one-third is blown out. Most of the lead which
deposits on the land is immobilized In the soil, but that which settles on
streets can be washed into storm sewers and ultimately into the coastal
waters during the rainy season. The lead vhich remains airborne is
primarily in submicron particles and can be transported yell beyond the
boundaries of the urban region. It must be borne in mind that large
uncertainties do exist for both the input and output terms for the lead
mass balance reported here.
THE LEAD IN THE SALINE BRANCH WATERSHED OF ILLINOIS
Another lead study was conducted by Kolfe and Haney (1975) for the
Saline Branch Watershed. This watershed includes the Champaign-Urbana
metropolitan area. The input of lead due to automobiles was determined
according to the number of vehicle-miles traveled. Although a complete
mass balance was not attempted, the ultimate fate of some of the lead
entering the watershed was determined.
Gasoline consumption data were used in conjunction with estimates of
the lead content Of gasoline to arrive at 29 tons/year emitted. It has been
estimated that 46 percent of the emitted lead remains airborne and is
transported by winds out of the watershed. Thus 16 tons/year is deposited
within the watershed. Of this, 14.3 tons/year is deposited either within
the urban area which accounts for 12 square miles (19 square km) out of a
total watershed area of 87 square miles (140 square km) or along major
highways.
As a check on the above lead input calculations, duatfall measurements
were made to determine the lead deposition at several locations in the
watershed. The result was 9.5 tons/year as compared with the 16 tons/year _
estimated above. It is argued that the estimated value (16 tons/year) Is ,
more accurate because the dustfall measurements probably excluded some of
the large material falling directly onto the roadway. ,
70
The only two significant exit pathways for lead from the watershed are
through the air or via drainage waters. The air pathway 'has not been
studied due to its complexity. The drainage waters have been examined,
however, and the results show that only 7.5 percent of the 16 tons/year
of lead deposited leaves the watershed. Most of the lead apparently
accumulates within the watershed.
The major reservoirs of lead appear to be soil, vegetation, and stream
sediments. Studies of animal life in the watershed show that a negligible
fraction of the lead is contained in the wildlife. Although a mass
balance has not been constructed, It has been demonstrated that a large
fraction of the lead entering the watershed remains within it.
ATMOSPHERIC TRACE-METAL FLOWS IN THE LOS ANGELES BASIN: ZINC AND CADMIUM
In this example of mass balances, the methods developed for the
analysis of the environmental flows of lead (Huntzicker et al., 1975a) are
applied to zinc and cadmium. The emissions of each element exhausted into
the atmosphere are quantified and the fate of these is traced through the
environment. Estimates are made of the amounts of each element depositing
over the land area of the Los Angeles Basin, advected out of the Basin, and
entering the coastal waters. These estimates have Important implications
for control strategies. In the case of inputs to coastal waters the most
important input route can be Identified for each element. The amounts of
these elements blown out of Los Angeles can be treated as primary inputs
in emissions inventories for the downwind regions.
Trace Element Deposition Measurements
The experimental aspects of this work have been discussed in detail by
Huntzicker et al. (1975a). Trace element deposition was determined on
specially designed Teflon substrates, and particle size distributions were
obtained with Andersen cascade impactors. Elemental analysis was by flame
atomic absorption spectroscopy, although flameless, carbon rod 'atomic
absorption was used in the later stages of the work,
71
-------
Depostion of zinc and cadmium was measured in Pasadena for several
periods between November 1972 and February 1974. A 1-week, synoptic
measurement of deposition at six sites (including Pasadena) around the
basin was made in July 1973. Deposition measurements were also made at
four sites along the coast, and on Santa Catalina and San Clemente Islands.
Deposition in Pasadena was simultaneously measured during the coastal
measurement.
Atmospheric concentrations of zinc and cadmium were obtained during
the 1972 and 1973 phases of the California Air Resources Board Aerosol
Characterization Experiment (ACHEX) (Hidy, 1973). Carbon monoxide data
were also obtained during ACHEX with Beckman GC 6800 air quality
chromatographs. All uncertainties cited in this paper represent one
standard deviation about the mean.
Zinc Emissions and Distribution
The major sources of atmospheric zinc In the Los Angeles area are
metallurgical emissions, tire dust, and automobile exhaust. Metallurgical
operations which are significant zinc emitters include secondary zinc
melting, zinc sweating, zinc vaporization (distillation), brass melting,
and Iron and steel melting. In general, zinc is emitted as the oxide.
The percentages of zinc In the emissions from the different sources
are summarized in Table 10. Total partlculate emissions are about 450
kg/day from zinc operations in Los Angeles County and about 900 kg/day
from brass melting (Thomas, 1974)*. The data In Table 10 show that the
zinc content of these emissions is in the range 46 to 78 percent, which
corresponds to zinc emissions of 600 to 1,100 kg/day. Total particulate
emissions from iron and steel manufacturing in Los Angeles County are
about 600 kg/day (Thomas, 1974)*. From the reported zinc composition of
Thomas, G. Los Angeles County Air Pollution Control District, private
communication (1974).
72
such emissions (Table 10), a zinc emission rate of 40 to 200 kg/day is
obtained. Thus total metallurgical zinc emissions in Los Angeles County
are in the range 600 to 1,300 kg/day. Orange County metallurgical
emissions of all types are only about 7 percent of the Los Angeles County
metallurgical emissions (State of California, 1972) and will not be
included in this analysis.
TABLE 10. ZINC CONTENT IN PARTICULATE EMISSIONS
FROM METALLURGICAL OPERATIONS
Source
Zinc sweating
Red and yellow brass
furnace (Los Angeles)
Brass furnace
Open hearth steel furnace
Electric arc steel furnace
Gray iron cupola furnace
(baghouse sample)
% Zn in Particulate
Emissions
46
47
<78
10-15
30
6
Reference
Herring (1971)
Allen et al. (1952)
Danielson (1973)
Danielson (1973)
Danielson (1973)
Danielson (1973)
The zinc oxide aerosol which is formed during evaporation from molten
zinc Is generally in very small particles (< 0.3 pm) (Herring, 1971).
Significant agglomeration apparently occurs, however, and the median
diameter of the effluent zinc oxide is about 1 pm (Herring, 1971).
Particulate matter emitted by steel melting is primarily CWO percent by
weight) in the 0.10-|Jm size range but, in iron casting operations, only
about 30 percent by weight of partlculate effluent is in this size range
(Danielson, 1973). It is not known, however, if the zinc particle size-
distribution is the same as the total mass distribution.
Tire dust in the atmosphere has been studied in detail by Peirson and
Brachaczek (1974). Pierson (1974) estimated that about 7.8 x 10
73
.-4
kg of
-------
tire dust Is produced .per-vehicle mile or 1.1 x 10~ kilogram per tire
mile. (The "per vehicle mile" figure includes a 20 percent contribution
fron 18-wheel trucks.) Extrapolating 1969 traffic data by a 3 percent
0
annual increase, approximately 1.3 x 10 miles per day was driven in the
Los Angeles area In 1973 (Roberts et al., 1971). Thus, production of
10 kg/day of tire dust is estimated. Pierson and Brachaczek (1974)
measured an average zinc content of 1.0 percent by weight for several
different tire treads. The resultant production rate of tire dust is
in very large particles (> 10 ym) with short atmospheric residence
times (Pierson and Brachaczek, 1974; Dannis, 1974a; Raybold and Byerly,
1972).
Another major source of atmospheric zinc is automobile exhaust. This
zinc most likely arises from the combustion of detergent lubricating oils
which contain about 0.1 percent of zinc by weight (Fierson and Brachaczek,
1974; Ondov et al., 1974). Cahill and Feeney (1973) sampled freeway aerosols
at several locations in Los Angeles and found that for particles less than
5 urn in aerodynamic diameter, the zinc-to-lead ratio is 0.013. This
particle-size range excludes most of the tire dust zinc. Ondov et al.
(1974) found a zinc-to-lead ratio of 0.012 for a tunnel study. Although
size-distribution data were not given, this ratio probably corresponds
to about the same size cutoff as the Cahill and Feeney ratio because of the
2-km length of the tunnel. The production rate of automobile exhaust
zinc can be estimated from the Zn/Pb ratios given above, the emission rate
(16,700 kg/day) of automobile exhaust partlculate lead In Los Angeles
(Huntzicker et al., 1975a) and the fraction of automobile exhaust lead in
particles below the tire dust range. For reasons discussed below, 7 ym
is the limit selected, although a cutoff of 5 ym would not significantly
affect the result. For 40 percent of automobile exhaust lead In particles
smaller than 7 \as (Habibi, 1973), the production rate of automobile exhaust
zinc is 80 kg/day. Ninomiya et al. (1971) measured zinc automobile exhaust
.emissions of about 400 mlcrograms per mile averaged over ten Federal Test
Procedure cycles. This corresponds to an emission rate of 50 kg/day for
the Los Angeles area. Current measurements provide an estimate of the
automobile exhaust zinc emission rate of 110 kg/day. This latter estimate
74
is based on a freeway zinc concentration measurement of 0.48 yg/ro of which
0.37 yg/m is attributable to automobile exhaust (dp < 7 ym), and lead
concentrations of 14.1 yg/m of which 12.0 yg/m Is automobile exhaust.
Background concentration measurements from this study and from 1972-1973
3 3
ACHEX data indicate that 0.17 yg/m of zinc and 2.2 yg/m of lead are the
background levels of automobile exhaust. Adjusting for these background
levels, the ratio of zinc to lead in automobile exhaust at the freeway is
0.017. Applying this ratio to the 6.7 tons/day of the automobile emission
estimate in the less than 7 ym range (Huntzicker et al., 1975a, and Habibi,
1973), the automobile zinc emission is calculated to be 110 kg/day. The
average of the above three automobile zinc emission estimates is 80 kg/day.
The total atmospheric emissions of zinc in the Los Angeles Basin are
1,700 to 2,400 kg/day. Inputs from other sources such as the weathering
of galvanized structures are presumed to be minor but are not known.
The fate of these atmospheric zinc emissions is removal by three major
environmental pathways: "near" (source) deposition, "far" or basin-wide
deposition, and advectlve (wind) removal.
To estimate the amount of airborne lead removed by wind from the Los
Angeles Basin, carbon monixide was used as a mass tracer for airborne lead.
While such an approach was acceptable for lead because lead and carbon
monoxide are emitted from the same source, this method is less satisfactory
for other species. The correct approach would involve a detailed, 3-
dlmensional mapping of pollutant concentrations and wind velocities as a
function of time. Lacking such data, an estimate can be made using the
carbon monoxide tracer method. In this application, carbon monoxide is
not to be considered a tracer for zinc and other trace-metal pollutants, but
rather a parameter which scales to the volumetric air flow through the
Basin.
Values for zinc-to-carbon monoxide ratio (Zn/CO) at several sites in
the Basin are given In Table 11; Zn/CO is 0.088 — 0.041 ym/m /ppm or
(7.6 — 3.6) x 10 as a dlmensionless weight ratio. The Los Angeles
75
-------
County Air Pollution Control District (MacBeth, 1974)* estimates that 6,450
metric tons/day of carbon monoxide was produced In the County by automobiles
in 1973. Orange County adds another 20 percent (State of California, 1972)
to bring the motor vehicle total to 7,700 tons/day. Stationary sources add
an additional 4 percent (State of California, 1972) for a grand total of
8,000 tons/day for 1973. From Zn/CO we estimate that 600 — 500 kg/day of
zinc is blown out of the Basin.
TABLE 11. ZINC TO CO AND CADMIUM TO CO RATIOS IN THE tOS ANGELES BASIN
Site
Pasadena
West Covina
Rubidoux
Dominguez Hills
Pomona
ZN/CO
g/m3/ppm
0.032
0.089
0.089
0.15
0.076
Cd/CO
0.002
0.002
0.004
0.003
0.001
Average •
0.087 + 0.041
0.002 + 0.001
The near deposition of zinc can be estimated from source and receptor
particle size-distributions as was the case for lead. Particle size-
distributions at a Los Angeles freeway and in Pasadena, a typical receptor
site, are shown in Figure 12. The dashed lines assume that the smallest
particle size is 0.01 ym and the largest 50 pm. The freeway distribution
shows considerably more mass above 7 microns than the Pasadena data.
This shows that much of the airborne tiredust zinc deposits near the
source. It should be cautioned, however, that these impactor runs were
nonisokinetic and, therefore, the concentration of large particles is
underestimated. It is still possible, however, to use the data to determine
an upper limit for near deposition.
MacBeth, W., Los Angeles County Air Pollution Control District,
private communication (1974).
76
0.4
0.2
Am/nlt 0.8
A log dp
0.6
0.4
0.2
X
FREEWAY Zn
\
.01
2/74
PASADENA Zn
0.1
1.0
dp (pi)
10
Figure 12. Differential mass distributions for zinc at 1 meter
from a freeway and for Pasadena.
According to the Pasadena distribution, about 14 percent of the zinc
mass is greater than 7 microns. If this distribution is typical of zinc
particle size-distributions around the Basin, then about 80 kg/day
(14 percent of 600 kg) is blown away from the roadway. The remainder of
the tire dust zinc, 920 kg/day, is assumed to have a short atmospheric
residence time and to deposit on or near the roadways. Because 80 kg/day
is a lower limit for zinc blown away from the roadway, 920 kg/day is
an upper limit for near deposition.
77
-------
The deposition of zinc as a function of distance from a Los Angeles
freeway (62,000 vehicles per day) IB shown in Figure 13. The steep
gradient near the roadway edge supports the concept of near deposition.
Similar gradients in soil zinc near highways have been found by Lagerwerff
and Specht (1970). Integration of deposition data from the roadway edge
to the final sampling point at 115 m gave only 10 percent of the expected
deposition estimated in the above discussion on near deposition. The
roadway at the sampling site was straight and slightly downhill, and it
is likely that tire wear at this site was less than average. It is- also
rosslble that a substantial amount of the tire dust remained on the
roadway.
Near deposition can also be expected for zinc emissions from the
ferrous metal industry. If the zinc particle size-distribution is assumed
to be the same as the total mass distribution, then about 50 percent of
the zinc is in particles greater than 10 )Jm. The estimated near deposition
of zinc Is then about 60 kg/day.
Measurements of zinc deposition at Pasadena and other receptor site
locations around the Basin can be used to evaluate the far deposition
component. The average Pasadena deposition for ten periods between
November 1972 and February 1974 was 14+3 ng/cm2/day. For the basin-
wide synoptic measurement, the average deposition normalized to Pasadena
was 1.1 + 0.6. This gives an average zinc deposition flux for the Los
2 2
Angeles Basin of 15+9 ng/cm /day. For an area of 4,430 km the basin-
wide deposition is 660 + 390 kg/day. This probably represents a lower
limit for surface deposition because it is based on the superficial area
of the Basin. More data are needed to evaluate the effects of vegetation
and man-made structures.
2.0
1.0
0.2
0.1
0.0004
0.0002
Pb
Zn
Cd
JL
_L
50 100
DISTANCE FROM FREEWAY |m)
A mass balance for Los Angeles zinc is shown in Table 12. The output
flows are for dry weather which accounts for more than 90 percent of"
the days in Los Angeles. Although the overall agreement is satisfactory
the limits of error are large, and it is possible that unidentified
Figure 13. Deposition as a function of distance from
freeway for Pb, Zn, and Cd.
78
79
-------
sources exist. The environmental significance of such sources can only
be detensined if their characteristics are known.
TABLE 12. MASS BALANCE FOR ZINC IN THE ATMOSPHERE
T8
Input (kg/day
Metallurgical 600-1300
Tire dust 1000
Auto exhaust 80
1700 - 2400
Output (kg/day)
Near deposition
Far deposition
Removed by wind
£ 980
660 +
600 +
•\, 2200
390
500
As was the case for lead, the output routes shown in Table 12 are not
final sinks. Some of the zinc which is blown out of the Basin can deposit on
the coastal waters. Zinc which has deposited on the streets and the land
can be washed into storm sewers and out to sea during rainstorms. These
environmental pathways are included in a diagram of the flow of zinc
through the Los Angeles Basin shown in Figure 14. The flows into coastal
waters are discussed later in this Section.
Cadmium Emissions and Distribution
Cadmium is generally considered to be a toxic substance, and conse-
quently a knowledge of its environmental flow is important. Its toxic
properties and broad-scale environmental flow have been reviewed in
considerable detail (Fulkerson and Goeller, 1973; Athanassiadis, 1969;
Friberg et al., 1971). However, little is known about the flew of cadmium
through specific urban environments. The lack of detailed emission
inventories has been the primary obstacle. The Los Angeles area is no
exception.
80
-------
The secondary zinc processing Industry is a source of atmospheric
(Fulkerson and Goeller, 1973), but emission rates are not well
knoan. Herring (1971) reports that the cadmium- to-zinc (Cd/Zn) ratio
for ^articulate emissions from a zinc— sweating reverbertory furnace was
abo-.t 0.0006. If this ratio can be applied to all of the secondary zinc
processing operations In Los Angeles, then cadmium emissions would be about
0.6 *g/day. Cd/Zn ratios are higher In effluents from primary zinc
processing operations, but there are apparently no such plants in Los
Ar.g<=Jes. Secondary operations use previously refined zinc, which is low
in csdmlura (Fulkerson and Goeller, 1973) and probably emit only small
anx>ir,ts of cadmium as indicated above. However, one of the zinc-smelting
plai-ts in Los Angeles also manufactures cadmium ball anodes, and it is
possible that this could be a source of atmospheric cadmium. The particulate
cadBlum emissions from this operation are unknown. There are also a number
of plants producing cadmium chemicals In the area, but their emissions
ar<= also unknown.
Lee and von Lehmden (1973) report that baghouse dust in brass and
bronze smelters contains between 100 and 10,000 ppm cadmium. Since 910
kg/day of particulate matter is emitted by brass melting operations in
Los Angeles, these fractions Imply cadmium emissions In the range
0.09 to 9 kg/day.
There are also a number of lead smelting and fabricating plants in
the (OB Angeles area. The Los Angeles County Air Pollution Control District
(Thomas, 1974)* reports daily emissions of about 300 kg/day for these plants,
but l,ee and von Lehmden (1973) report an upper limit of 1 ppm cadmium in
particulate emissions from such sources. This corresponds to less than
0.3 kg/day.
Another possible source of cadmium is the automobile. The ratio of
cadmtum to zinc in tire'dust is about 0.0006 (Fulkerson and Goeller, 1973;
Thomas, G., Los Angeles County Air Pollution Control District, private
communication (1974).
82
Plerson, 1974)* which corresponds to an emission rate of 0.6 kg/day. As
noted above, most of this will deposit on or near roadways. An analysis
of three motor oils and twelve brands of gasoline gave cadmium concentrations
of about 0.2 ppm for the former and less than 0.01 ppm for the latter
(Lagerwerff and Specht, 1970). The resultant automotive cadmium emission
Is less than 1 kg/day.
All of these emission sources add up to a few kilograms per day
at most. The Los Angeles County Air Pollution Control District estimates
an emission rate of less than 40 kg/day. As will be seen below, at least
20 kg/day is necessary to satisfy the measured environmental flows.
Values of cadmium to carbon monoxide ratio (Cd/CO) at several sites
in the Basin are given in Table 11, From the average value of Cd/CO "
0.002 - 0.001 vg/m /ppm, an estimate of 14 - 10 kg/day of cadmium is
advected from the Basin by wind. The average deposition flux for five
different measurement periods in Pasadena was 0.13 - 0.07 ng/on /day.
Because cadmium data were obtained during the urban synoptic measurements,
basin-wide (far) deposition is based on the Pasadena deposition. This
estimate is 6 - 3 kg/day.
The deposition of cadmium as a function of distance from a freeway
is shown in Figure 13. Similar gradients for cadmium in roadside soil
and vegetation have been reported by Lagerwerff and Specht (1970). The
gradient is much weaker than the corresponding zinc profile, but a
factor of two difference, between the 1 m and 115 m points, still exists.
Integration between the roadway and the 115 m point (assuming that the
latter point represents "background") gives an emission rate of about
3 yg per car mile or 0.4 kg/day for the whole Basin. This is almost
70 percent of the amount expected from tire dust. Because zinc deposition
at the same location account for only about 10 percent of the amount
expected from tire dust, the existence of other automotive sources may be
Indicated.
Pierson,_ D.H., Private communication (1974).
83
-------
The Input and output flows are summarized in Table 13. The known
sources account for (at most) one-half of the environmental flow and it is
Impossible to judge whether all of the cadmium emitted into the atmosphere
has been accounted for. Clearly better emissions data are required.
TABLE 13. MASS BALANCE FOR CADMIUM IN THE ATMOSPHERE
Input (kg/day)
Secondary zinc processing
Cadmium processing
Brass melting
Lead smelting
Tire dust
Auto exhaust
0.6
0.09 to 9
0.3
0.6
< 1.0
3-11
Output (kg/day)
Near (roadway) deposition 0.4
Far deposition 6 + 3
Removed by wind 14 + 10
20+13
Trace-Metal Inputs to the Coastal Waters
Trace metals can enter the coastal waters by at least six routes:
dry atmospheric deposition, rainout-washout, rainy season runoff, dry
season runoff, discharge of treated wastewaters, and discharge of untreated
wastes. Estimates based on available information for all of these routes
can be made except for the discharge of untreated wastes. Analyses will
depend heavily on work discussed earlier on automobile-emitted lead to
the coastal waters (Huntzicker et al., 1975b).
The deposition of lead, zinc, and cadmium normalized to their
respective coastal values are plotted as a function of distance from the
coastal baseline in Figure 15. The solid line is an extrapolation to
the coast, and the dashed line represents "background" or contribution from
non-Los Angeles lead deposition (Huntzicker et al., 1975b). The open
symbols represent 1973; filled symbols represent 1974. The average
84
f\
coastal deposition values are: lead, 10 ng/cm /day, zinc, 7.6
2 7
ng/cm /day and cadmium, 0.12 ng/cm /day. These values are based on
measurements conducted for this analysis and do not average in deposi-
tion flux data reported by others.
Zn/Pb ratios for several locations are listed in Table 14. The
difference between the inland air and inland deposition ratios can be
attributed to different particle size-distributions for lead and zinc. Large
enrichments of zinc relative to lead are observed for Catalina rain
(Lazrus, et al., 1970) and for sediments in Santa Barbara, Santa Monica,
and San Pedro marine basins (Bruland et al., 1974). Lazrus1 data were
50 "100
DISTANCE FROM COASTAL BASELINE (km)
Figure 15. Deposition of Pb, Zn, and Cd as a function of distance
from the coast.
85
-------
obtained at the Catalina airport weather station which is near a zinc
nine; this may have Influenced his data. The sediments are probably
Influenced by the large amounts of zinc present In the sewage effluents
and storm water runoff which are discharged Into the coastal waters.
Smaller enrichments of zinc are seen In coastal deposition with
respect to Inland deposition and in San Nicolas Island air with respect
to inland air (Hidy et al., 1974). The enrichment in coastal deposition
cannot be attributed to zinc in windblown sand because an insufficient
amount of sand on the deposition collector was observed. It is possible
that Industrial sources located along the coast are responsible. In
Table IS the ratios of coastal deposition to inland deposition and Catalina
deposition to inland deposition are given for lead, zinc, and cadmium.
Although the uncertainties are large, the smallest ratio in both cases was
TABLE 14. Zn/Pb AT VARIOUS LOCATIONS IN THE LOS ANGELES AREA
Airborne
Los Angeles air (Inland)
Santa Catalina Is. rain (Lazrus et al., 1970)
San Nicolas Is. air (Hidy et al., 1974)
Deposition
Los Angeles urban (inland)
Los Angeles coastal
Santa Catalina Is. (area-of high zinc soils, August 1973)
Santa Catalina Is. (relocated sample, October 1974)
San Clemente Is.
Marine
Storm runoff (December 21, 22, 1972)
Zn/Pb Ratio
0.12
2.60
0.50
0.33
0.75
5.50
0.85
0.25
1.4
(Continued)
86
TABLE-14. Zn/Pb AT VARIOUS LOCAIIOHS IH THE LOS AHGELES AEEA (Continued)
Airborne
Dry weather runoff (July 1973)
Sediment accumulation rates (Bruland et al., 1974)
(natural)*
(anthropogenic)
Sewage effluent
Miscellaneous
Santa Monica sand (25 ppm Zn)
determined from pre-1920 sediments
Zn/Pb Ratio
0.7
3.9
11.0
1.5
5.7
1.6
observed for lead for which most of the emissions are inland. Larger ratios
were observed for zinc and cadmium; this is consistent with the fact that
some stationary sources are located along the coast.
Coastal
Pb
Zn
Cd
0.
0.
0.
2 H
5 H
9 H
Inland
l- 0.
t °-
t °-
1
4
7
Santa
0.07
0.80
0.22
Catalina Inland
(Oct.
(Oct.
1974
1974
0.
0.
008)
02)
The factor, of-9 decrease, in Catalina, to inland lead- deposition
between the August 1,973 and October 1974 experiments may have resulted
from differences in meteorology during the two experiments. However, the
equivalent ratio for zinc shows a factor of 40 decrease. This may have
been caused by changes in meteorology and by the lower zinc content of
the soil at the location of the October 1974 run.
87
-------
The atnospherlc deposition of cadmium on the coastal waters is
estiaated by assuming that the deposition decreases linearly between
the coast and Catalina (solid straight line in Figure 15). This defines
an area of influence of Los Angeles which extends to about 70 km from the
2
shore and gives an area of 5600 km for an 80 km long coastline. The
resultant cadmium deposition flux to this coastal region is 3.5 kg per
day. The Input of Los Angeles zinc to the coastal waters is assumed to
fall along the solid line in Figure 15. The resultant flux is 200 kg/day.
No corrections for non-Los Angeles background were made as In the lead
analysis. On an annual basis,* the inputs are 70 tons/year of zinc and
1 ton/year of cadmium.
The rainfall fluxes of Los Angeles zinc and cadmium are estimated
from the rainfall flux of lead (30 tons/year) and the ratios of airborne
Zn/Pb and Cd/Pb (0.12 and 0.004, respectively) in Los Angeles air. Thus
approximately 4 tons/year of zinc and 0.1 tons/year of cadmium enter this
coastal water region by rainout-washout. As noted earlier the region of
influence of Los Angeles with respect to rainout-washout may be larger than
the region of influence for dry deposition, but to confirm this requires
additional data that are not currently available.
To estimate the amount of lead entering the coastal waters by runoff
in the rainy season, Huntzicker et al. (1975b) developed a simple model
based on measured runoff fluxes for a storm on December 21, 1971 and
several assumptions concerning the accumulation and removal rates of lead
in street dirt. By scaling the model-dependent lead flux (140 tons/year)
by the ratios Zn/Pb (1.4) and Cd/Pb (0.01), for the December 21, 1971
storm, zinc and cadmium runoff fluxes can be estimated. These runoff
fluxes are 170 tons/year for zinc and 1 ton/year for cadmium. As was the
case for lead, more data are needed to better define these fluxes.
There is an average of 19 days per year in the Los Angeles region in
which the total rainfall is greater than 2.5 mm (U.S. Department of
Commerce, 1968-1973). A "dry" day is defined as one for which the total
rainfall was less than 2.5 mm, and the coastal water deposition fluxes
are calculated on the basis of 346 dry days per year.
88
The zinc runoff of about 500 kg/day is more than half of the estimated
near deposition. The application of the lead runoff model to zinc implied
an accumulation rate of zinc in street dirt of more than twice the estimate
for near deposition of zinc. However, Pitt and Amy (1973) found that the
average loading of zinc In street dirt for several American cities was
only about one-fourth of the loading of lead. Thus, a substantial
fraction of runoff zinc probably comes from non-street dirt, nonautomotive
sources (e.g., corrosion of galvanized structures).
The trace-metal inputs from runoff during the dry season can be
estimated from runoff fluxes In the Ballona Creek drainage basin. From a
sample taken in August 1973 the estimated flows to the drainage basin are
2.4 kg/day for zinc and 0.1 kg/day for cadmium. The Ballona Creek drainage
basin constitutes 5 percent of the area of the Los Angeles study region.
An extrapolation on the basis of area to the whole region gives dry runoff
fluxes of 20 tons/year for zinc and 0.7 tons/year for cadmium.
Trace metals also enter the coastal waters in sewage effluents. These
fluxes are 1,300 tons/year of zinc and 27 tons/year of cadmium (Parkhurst,
1973; City of Los Angeles, 1973; and County Sanitation Districts of Orange
County, 1973). In Table 16 the trace-metal flows for all the environmental
pathways into the coastal water are summarized. Despite substantial
uncertainty in these estimates, it is clear that sewage effluent is the
dominant input for zinc and cadmium, and that control strategies for these
elements should focus on the sewage route. In the case of lead, however,
the "atmospheric" routes (including runoff) are important and must be
considered in a coastal water lead-control strategy.
89
-------
-II •-'•"*--
TABLE 16. ANTHROPOGENIC INPUTS OF LOS ANGELES Pb, Zn, AND Cd
Input Route
Storm runoff
Dry deposition
Rainout-vashout
Dry runoff
Municipal sewage
Pb
140
120
30
10
230
Tons /Year
Zn
170
70
4
20
1300
Cd
1
1
1
0.07
27
90
SECTION V
TRANSPORT AND DEPOSITION THROUGH THE WATER ROUTE
FLOWS OF ARSENIC IN THE SOUTHERN CALIFORNIA WATER ENVIRONMENT
Arsenic eaters the Southern California coastal basin in stream flows,
ground waters, water supply aqueducts, atmospheric transport, and in the
form of raw materials and manufactured goods. It is discharged to the
coastal waters in stream flows and in domestic-industrial wastewater
effluents (liquid and solid). Arsenic accumulates in coastal sediments in
the vicinity of wastewater discharges. Arsenic is also added to the
land as sludges recovered from municipal wastewater processing. Atmospheric
emissions within the basin may be deposited on the land or on the coastal
waters, or transported from the basin by prevailing winds. However,
recent information on arsenic emissions to the atmosphere or on ambient air
concentrations of arsenic in the Los Angeles Basin have not been
identified.
A preliminary assessment of arsenic flows In the Southern California
water environment can be made through an examination of water quality and
hydrologic data of the U.S. Geological Survey, influent and effluent
wastewater characteristics reported by the County Sanitation Districts of
Los Angeles anoVthe-City of Los Angeles, and reports of the Southern
California Coastal Water Research Project (SCCWRP) on mass emissions of
metals.
The mean annual runoff of rivers discharging to the Southern California
Bight is about 600 million m /year (SCCWRP, 1973). Mean daily runoff is
91
-------
highly variable during storm periods. On twelve dates in the period from
February 6, 1973, through October 20, 1975, the Geological Survey measured
arsenic in samples from the Los Angeles River at Long Beach (U.S. Geological
Survey, 1973). Total arsenic concentrations in the twelve samples ranged
from 3 to 26 yg/1; mean daily discharges for the twelve sampling dates
varied from 0.6 to 60 m /sec. The U. S. Geological Survey data reveal
that arsenic is present in both particulate and dissolved fractions in
most Southern California river water. Since the flow and arsenic concen-
tration are uncorrelated, the annual flux may be estimated as the product
of the annual flow of the Los Angeles River (150 x 10 m /year) and the mean
concentration (10 yg/1), or 1.5 tons/year.
Since the mean annual runoff of the Los Angeles River is about
25 percent of the total annual stream discharge to the bight, a rough
estimate of the total arsenic flux to the bight from all streams, based
on the Los Angeles River data, is about 6 tons/year.
An alternative approach is to select an average arsenic concentration
for the thirty individual analytical results for the three rivers
sampled—Los Angeles River, Santa Clara River, and Santa Ana River— and
to estimate a typical annual flux as the product of the average concentration
and the mean annual runoff for the entire Southern California Bight. The
mean of thirty total arsenic values for the three rivers (1973-75) is
7 yg/1. The typical river flux to the entire bight is then estimated
about 4 tons/year.
The discharge of municipal wastewaters to the Southern California
Bight totaled 1,431 million m3/year in 1974 (SCCWRP, 1975). The total
mass emission rate (flux) for arsenic from all municipal waste discharges
was estimated by SCCWRP to be greater than 21 tons/year in 1974. The
dominant discharges of arsenic are from the Los Angeles County Joint
Water Pollution Control Plant (JWPCP) outfalls off Palos Verde (11.3
tons/year), the City of Los Angeles (Hyperion) final effluent to Santa
Monica Bay (4.7 tons/year)., the sludge line to Santa Monica Canyon (1.2
tons/year), and the San Diego outfall at Point Loma (greater than 3.7
92
tons/year). Arsenic data have not been reported for the Orange County
outfall (flow of 236 million m3/year) .
Arsenic entering the Hyperion treatment plant in the 1971-1972 period
was estimated at 4.1 tons/year from domestic waste and about 1 ton/year
from Industrial waste (City of Los Angeles, 1973), or 81 percent from
domestic and 19 percent from industrial sources. These are the same
proportions for the wastewater flow. Measured concentrations did not
reflect quality differences between residential and industrial wastewaters
with respect to arsenic. The City of Los Angeles has identified industrial
processes which produce products which contribute to the arsenic mass
flux at Hyperion. These are flue dust of lead and copper smelting,
hardening lead shots, copper alloys, lead base alloys for battery grids,
bearings, cable sheaths, anodes, high-temperature brasses, copper alloy
tubes, dental amalgams, solders, and pesticides (City of Los Angeles, 1973).
The Sanitation Districts of Los Angeles County have identified the follow-
ing types of industry as emitting arsenic to the sewer system: blast
furnaces and mills, ferrous foundries, nonferrous foundries, and apparel
(City of Los Angeles, 1973). From the data reported by the City of Los
Angeles for 1971-1972, 7.1 tons/year arsenic is estimated in Hyperion plant
influent. This corresponds to a concentration of 15 yg/1 in a mean flow
of 344 million gallons per day (MGD) (475 million m /year). This estimate
is not very reliable because of the low number of analyses for arsenic.
The weighted-water-supply source contribution of arsenic to Hyperion
flow for 1971-1972 is estimated at 6.7 tons/year, based on treatment plant
flows and concentrations reported for the Owens River aqueduct, groundwater,
and Colorado River water (City of Los Angeles, 1973). Total arsenic
emissions from Hyperion (effluent plus sludge outfalls) for the years 1971,
1973, 1974, and 1975 have been estimated by SCCWRP to be 3.5, 7.4, 5.9,
and 4.8 tons/year, respectively (SCCWRP, 1974, 1975, 1976). There is reason
to believe that the 1971 data for arsenic are not reliable because of the
Insufficient sensitivity of earlier analytical methods (SCCWRP, 1974).
The arsenic content of sludge solids discharged through the Hyperion
7-mile outfall (arsenic flux, 1 to 2 tons/year) can be estimated from
93
-------
data on the total suspended solids concentration and total arsenic con-
centration of the sludge, assuming that all arsenic Is In the sludge
partlculates. The average arsenic concentration of the sludge in 1974
was 0.18 rag/1, and the average suspended solids concentration was
7,300 mg/1. The arsenic content of the sludge solids ia then estimated
at 25 mg/kg as an upper limit. Observations on arsenic contents of
bottom sediments in the vicinity of the 7-mlle outfall (City of Los
Angeles, 1973) show accumulations of perhaps 2 to 3 times background levels.
Arsenic flows at the JWFCF of the County Sanitation Districts of Los
Angeles (discharged to the ocean at White's Point) appear to have increased
from 1970 through 1974, but this may be only a reflection of the unreliability
of earlier analytical results for arsenic (data frequently reported as 0.01
mg/1). Arsenic flows at the JWPCP have been estimated as follows:
Flow
From Water Supply
(1970)
Influent (1971)
Effluent
(1973)
(1974)
(1975)
Ton/Year
< 4.5
< 5.3
7.4
11.3
5.2
Reference
Sanitation Districts of
Los Angeles County, 1973
Sanitation Districts of
Los Angeles County, 1973
SCCWRP, 1975
SCCWRP, 1974
SCCWRP, 1976
The 1974 JWPCP arsenic flow to the ocean represents a total arsenic
concentration of 25 yg/1 In a wastewater flow of 346 million gallons per
day (477 million n>3/year).
Arsenic concentrations in seawater are reported to range from 0.8
to 8.0 Pg/l, with a representative total concentration of 3.7 yg/1 (Brewer,
1975).
94
In these data, the advectlve transport of arsenic through the Southern
California Bight (surface area 1 x 1011 m2, mean depth of surface mixed
zone 50 m, residence time 3 months) is about 70,000 tons/year. The ratio
of total arsenic concentration in JWPCP effluent to typical arsenic concen-
tration in seawater is 25 to 3.7, or 6.8. Dilution of 1 part of effluent
with 100 parts of seawater results in an arsenic concentration in the
mixture of 3.9 Pg/l, or an increment over background of 0.2 Mg/1.
Typical concentration ranges for arsenic in waters, wastewaters, sludge
and sediments In Southern California are as follows:
Water Supplies
Stream Flows
Wastewater Effluents
Wastewater Sludges
Marine Sediments
(background)
Arsenic Concentrations
< 10 to 30 yg/1
1 to 26 yg/1
7 yg/1 (mean)
10 to 30 yg/1
150 to 300 yg/1 (sludge)
15 to 30 mg/kg (solids)
1 mg/kg
Marine Sediments & Sludge
(near outfalls)
4 to 9 mg/kg
Estimated typical arsenic total flows in the Southern California
environment are as follows:
Metric Tons Per Year
Streams
Wastewater Effluents
Wastewater Sludges
California Bight Advectlon
4 to 6 (1973-75)
18 to 20 (1974), 10 (1975)
1 to 2 (1974), 2 (1975)
70,000
95
-------
Arsenic flows in Southern California wascewater effluents for
1975 have recently been summarized by SCCWRP (1976). The total mass
emission of arsenic to the bight, exclusive of Orange County effluent, was
estimated to be 12 tons/year. A major decrease from 11.3 to 5.2 tons/year
in comparison to 1974 data is reported for the JWPCP (Los Angeles County)
effluent. The Point Loma discharge (San Diego) of arsenic decreased from
more than 3.73 to 0.15 tons/year. These changes probably reflect improved
analytical methods for arsenic. Source control programs may also have
influenced effluent levels of arsenic.
On the basis of this preliminary assessment it appears that wastewater
effluents are the dominant water transport influence in increased arsenic
fluxes to the coastal waters. With expected seawater-to-wastewater
dilution ratios on the order of 100:1, a concentration increase above
background in the water column of less than 5 percent should result.
Increases of 5-to 10-fold In arsenic content of coastal sediments are
experienced in the immediate vicinity of outfalls in Santa Monica Bay,
reflecting accumulation of arsenic-containing sludge solids and effluent
participates.
Data on atmospheric emissions and ambient air concentrations of arsenic
in Los Angeles are needed to complete the mass balance assessment of-arsenic
in the Los Angeles Basin.
THE ULTIMATE FATE OF SEWAGE PARTICULATE MATTER
Many studies have been conducted involving simultaneous diffusion
and sedimentation of particles during long-term transport from the source.
These studies have involved particles in both air and water.
Considerable meteorological data are available on wind and temperature
profiles, and on the nature of atmospheric turbulence. These data are
useful in validating the modeling of transport processes In the atmosphere.
A different situation exists for the long-term transport of particles in
bodies of water where current data are scarce. Of the many studies
96
involving particle behavior in oceans, for example, few have attempted to
incorporate actual ocean current data and turbulence characteristics into
the various models.
A study has been conducted concerning the physical behavior of
particles from ocean outfalls. The object of this study was to determine
where sewage or sludge particles eventually come to rest on the ocean
floor after being discharged from a submarine outfall diffuaer pipe. The
relevant parameters considered include the size, distribution, and densities
(hence settling velocities) of emitted particles, the ocean currents, the
structure of the ocean bottom, the temperature and salinity of the ocean
water, and the composition of the effluent sewage. Given appropriate
values for these parameters, it is possible to use the models developed by
Koh* to predict the eventual "fallout distributions" of the ocean bottom.
A brief summary of the methodology used in making these predictions is
presented here.
Typical submarine outfalls discharge treated effluent several
kilometers from shore through diffuser pipes on the ocean bottom, usually
at depths of about 60 m on the West Coast. The effluents contain some
fine particles not removed in the treatment. Furthermore, some digested
sludge particles may either be added back In the effluent stream or discharged
through a separate small sludge outfall.
After being discharged into the ocean, the wastewater plume rises
until reaching density equilibrium. The wastewater has a density slightly
less than the surrounding seawater. This density difference drives the
Initial plume rise. The temperature and salinity structure of the ocean,
as well as currents, then control the extent to which the plume rises.
Koh, R. C. Y., "Physical Disposition of Particulates from Ocean
Outfalls," Report to the Environmental Protection Agency, W. M. Keck
Laboratory of Hydraulics and Water Resources, Pasadena, California,
(in preparation).
97
-------
The plume rise occurs over relatively short time scales, on the
order of minutes or tens of minutes. Because the sedimentation velocities
of the sewage particles are small (10 to 10~ cm/sec), their fall
distance over the time scale of the plume rise is negligible compared to
that of the plume rise itself. It may therefore be assumed that the particle
settling does not begin until after the plume has found its equilibrium
position. The concentration distribution of sewage particles after the
plume rise has ceased represents the initial condition for the following
models.
Model 1; Differential Equation Approach
The idea behind this model Is to obtain a numerical solution to the
steady advective diffusion equation representing the transport processes
after the initial plume rise. Particle coagulation and exchange with the
surrounding fluid are not taken into account. The equation and boundary
conditions pose a boundary value problem which may be solved using a digital
computer. The basic equation for a collection of particles with
sedimentation velocity v is:
(Vy)
(vz) >
(17)
where the x-axls is parallel to the ocean current, y is the transverse
coordinate, and z is the vertical coordinate from the bottom in constant
depth H(0 _< z <_ H); K (z) and KZ(Z) are the eddy diffusivities for the
y-z-directions, respectively, and c(x,y, z) is the concentration of
particulate matter. It is assumed that the current speed u(z) is steady
and unidirectional, and that diffusion in the direction of current flow
is negligible. The boundary conditions are:
= H, x > 0
(18)
0, x > 0
c " CQ(y,z) given
98
Solving the above boundary value problem numrically would require
long computation times because of the need for Urge 3-dimensional
arrays. The method of moments can be applied to the system to reduce
the number of independent variables to two, thereby reducing computation
times. The disadvantage of this approach is the loss of some of the
detailed information about the concentration field.
t
The ± moment of c(x,y,z) is defined as:
t(x,z) = I y1 c(x,y,z) dy
Equations 17 and 18 can be multiplied by y1 and Integrated from -» to
(19)
•+<» to yield equations for the 1 moment.
The odd moments are zero by symmetry. Most of the Information about
c is contained in the zeroth and second moments. The zeroth moment
represents the number of particles per cm in the xz plane, while the
second moment provides a measure of the width of the plume in the
transverse (y) direction.
In the original paper, the moment equations are in finite difference
form and empirical functions for K (y) and K (z) are from the literature.
The equations were solved numerically, and an example of the results is
shown in Figure 16. The vertical dlstrlbi
for various times after particle release.
shown in Figure 16. The vertical distributions of c (x,z) are plotted
Model 2: Simulation Approach
One disadvantage of the diffusion equation model described above is
that a constant ocean depth must be assumed. With the simulation model,
bathymetry may be taken into account. In addition, the detailed structure
of the ocean current may be Incorporated into this model Instead of
assuming a steady unidirectional current.
99
-------
70
60
• T = OHRS
• T = 1.2 HRS
* T = 3HBS
O T = 9 HRS
X T = 34 HRS
SEDIMENTATION VELOCITY |VS| = 3 x lO^m/s
P 40
£ 30 -
0.4 0.6
REL. AMT
Figure 16.
Vertical distribution of particle concentrations at
the centerline as a function of travel time.
100
The basic idea behind the simulation approach is chat a particle
is carried from its initial location (x , yQ, z^) by longitudinal and
transverse ocean currents u and v while it settles with velocity vg. It
eventually reaches the bottom at point (x,y,z.), where z, is the height
of the ocean bottom at (x,y) above some reference height (such as the
greatest depth in the xy range of interest). All of the detailed
information about ocean currents is contained in u and v. The path of
the particle transported by u and w while settling at a velocity v can
be obtained using a digital computer.
The difficulty with this method is that detailed data on ocean
currents are lacking. It is therefore necessary to synthesize data by a
variety of methods, in addition to using the limited data available. An
example of synthetically generated ocean current data is shown in Figure
17.
Although the simulation approach can be adapted to include bathymetry,
it will be presented here only for the case of a constant ocean depth. A
discussion of modifying the procedure for variable depths is included in
the original paper.
The goal of this model is to obtain the bottom fallout pattern B(x,y)
at z = 0. This is accomplished by defining probability density functions
(pdf) for several variables and integrating these pdf's numerically.
Table 17 defines the various probabilities.
TABLE 17. PROBABILITY DEFINITIONS
g(z)dz
probability that a particle emitted by the
outfall is at an elevation between z and
(z + dz) after the dynamic plume behavior
has subsided
(Continued)
101
-------
TABLE 17. PBOBABItm DEFINITIONS (Continued)
EOT
f (v Jchr
fraction of particles with settling
velocities between v and (v + dv )
probability that a particle initially at
(xo,yo>zo) with settling velocity v
will settle to the bottom position (x,y)
KNOTS
* g £ £
h(t)dt
probability that a particle
outfall will have a fall
(t + dt)
The bottom fallout pattern resulting from a continuous
particles is given by:
re "•
B)
emitted by the c
time between ' o-
D
g. S
n
!
injection of i _
B(x.y) • /// / c(xo,yo,zo;x,y;vg) f(vs) dv g(z ) dzQS(xo,y )dx dyQ (20)
where s(x , y ) is the horizontal source distribution. It is assumed that
the density functions are mutually Independent in the above expression. If
S(xo,yQ) is taken as 3(xo>
Integral becomes:
Bg(x,y) - //f(wg) g(
J
It Is possible to combine
t • z/vs where h(t)
where h(t) is the density
becomes:
B3(x,y) - j
3(y ), where 3 is the delta function, the
z ) c.(z ; x,y;v ) dv dz
(21)
m
*<
P
r
CD
! 5
° i
PS i
I i
rt
S
rt
g-
rt M
o
the variables z and v into the fall time t:
8 3
= -i-2 / f(z/t)g(z)zdz,
function for fall time. Then the
QO
h(t) Cg (x.y.t) dt
(22)
». 1
to
o
expression 5*
i
I-1
(23) |
^V __jf
^3^ ->
" ? |
• ^^ ^-— ^
^^> " ^.
>\ ^_Jj
<^ /* =
"^ V "
J^^ f_
<^T ^r-
^ ^
^5 f
^^- *
i i i i
102
-------
where Cj(x,y,t) represents the probability that a particle will have a
horizontal displacement (x,y) in time t. The function Cg(x,y,t) can be
estimated from particle-release experimental or simulated data. If
particles are released Into the ocean at time t = 0, x « y = 0, and z > 0,
then the positions of the particles at time t. in the xy plane can be used
to determine Cg(x,y,t.). The bottom fallout pattern can be obtained as
follows:
1) From the available ocean current data, obtain the characteristics
necessary for the synthetic generation of ocean currents. Figure
17 shows such an example of synthesized data.
2) Determine h(t) from a knowledge of g(z) and f(v ).
3) Estimate Cj(x,y,t) for a particular t^, using transport data
(real or synthesized).
4) Repeat (3) as many times as desired for other values of t.
5) Sum the results with weight given by h(t) to obtain the fallout
pattern.
104
SECTION VI
CONCLUDING REMAKKS
The material balance-flow pathway approach presented in this report
is general in nature and can be applied to most environmental pollutants in
different urban areas. Although the method does not reveal the details
of pollutant dispersion in the environment, the requirement of a mass
balance demands that all important environmental pathways be identified
and quantified. Runoff and long distance transport will affect the
Inflows of pollutants downstream or downwind of the boundaries of an urban
region. Thus, the control of emissions In one region can materially
affect inflows into adjacent regions. This accounting method, then,
is potentially a powerful tool in assessing the environmental impact of
a pollutant. Unfortunately, however, large uncertainties will exist for
both input and output terms in the mass balance because of deficiencies
in data bases. Refinement of this approach will require more detailed
source characterizations and a better theoretical and experimental
understanding of particle removal processes in the atmosphere.
105
-------
r
SECTION VII
REFERENCES
Allen, G. L., Viets, F. H., and McCabe, L. C., Bureau of Mines Information
Information Circular 7627, O.S. Dept. of Interior, Washington,
D.C. (1952).
Anchanassladls, 7. C., "Air Pollution Aspects of Cadmium and Its
Compounds," NAPCA Report FB 188 086, APTD 69-32 (1969).
Atkins, P. R., J.APCA. 19, 591 (1969).
Boven, D. H. M., "Waste Lube Oils Pose Disposal Dilemma," Environ.
Sci. Technol.. j6, 25 (1972).
Brewer, P. 6., "Minor Elements in Seawater," In Chemical Oceanography.
J. P. Riley and G. Skirrow, eds. _t (1975).
Brief, R. S., Arch. Env. Health. _5, 527 (1962).
Bruland, K. W.. Bertlne. K., Koide, M., and Goldberg, E. D., Environ.
Sci. Technol. 8, 425 (1974).
Bullas, D. 0., 1953, data published by Chamberlain, A. C. (1960).
Bullock, J., and Lewis, W. H., Atmos. Env., 2, 517 (1968).
Cahlll, T. A. and Feeney, P. J., "Contribution of Freeway Traffic
to Airborne Partlculate Hatter," University of California, Davis,
UCD-CNL 169, Final Report to the California Air Resources Board
(1973).
California State Board of Equalization, Motor Vehicle Fuel (Gasoline)
Distributions. Department of Business Taxes (1972).
Cannon, H. L., and Bowles, J. M., Science. 137. 765 (1962).
Cardina, J. A., Rubber Chem. and Tech.. ^7 1005 (1974).
106
r
REFERENCES (Continued)
Carey, R. T. (Ed.), "California County Fact Book 1973," (County
Supervisors Association of California, California County Government
Education Foundation, 1973). The actual percentage of California
vehicles in the metropolitan parts of Los Angeles and Orange Counties
Is somewhat less than 41.4 because not all of the tvo counties are
Included in the study region. The excluded areas are sparsely
populated, however, and do not contribute significantly to the
automobile population.
Chamberlain, A. C., "Aspects of Deposition of Radioactive and other Gases
and Particles," in Aerodynamic Capture of Particles. E. G. Richardson,
ed, Permagon Press, New York (1960).
Chamberlain, A. C., Proc. Roy. Soc.. 296, 45 (1966).
Chamberlain, A. C., Proc. Roy. Soc. 296. 45 (1967a).
ChamBerlain, A. C., Con temp. Phys., JS, 561 (1967b).
Chamberlain, A. C., "Travel and Deposition of Lead Aerosols," Environmental
and Medical Sciences Division, UKAEA Research Group, Harwell
(February 1974).
Cholak, J., Schafer, L. J. and Teager, D., Amer. Ind. Hyg. Assoc. J., 29,
562 (1968).
Chow, T. J., Nature. 225, 295 (1970a).
Chow, T. J., "Environmental Pollution from Industrial Lead," International
Conference on Geochemistry and Hydrochemistry, Tokyo, (1970b).
Chow, T. J., Burland, K. W., Bertlne, K. K., Soutar, A. Kolde, M.,
Goldberg, E. D., Science. 181. 551 (1973).
City of Los Angeles, Department of Public Works, "Technical Report:
Waste Discharges to the Ocean" (January 1973).
Coluccl, J. M., Begeman, C. R., and Kumler, K., J. APCA, 19, 255 (1969).
County Sanitation Districts of Orange County, "Technical Report for
Water Quality Control Plan Ocean Waters of California" (January 1973).
Dalnes, R. H., Motto, H., and Chllko, D. M., Environ. Set. Technol.. £,
318 (1970).
Danlelson, J. A., Ed., "Air Pollution Engineering Manual" EPA Report
4 AP-40, Second Edition (1973).
107
-------
Dannis, M! L., Symposium on Ecology, ACS Rubber Division, Toronto,
Ontario, May (1974a).
Dannis, M.! L. , Rubber Chem. and Tech. . ^7, 1011 (1974b).
Davidson, D. I., Huntzicker, J. J. , and Friedlander, S. K., "The Flow
of Trace Elements Through tne Los Angeles Basin: Impact on Nonurban
Aread," Division of Enviromental Chemistry, 167th Meeting, ACS,
Los J^ngeles, Calif. (March 1974).
i
Dedolph, R. , Ter Haar, G. , Holtzmann, R., and Lucas, H. , Jr., Environ.
Sci.!Technol. 4^ 217 (1970).
DeMarrais.G. A., Holtzworth, G. C. , and Hosier, C. R., "Meteorological
Summaries Pertinent to Atmospheric Transport and Dispersion over
Southern California," Technical Paper No. 54, U.S. Weather Bureau,
Washington, D.C.(1965).
Federal Register, "Control of Mr Pollution from New Motor Vehicles and
New Motor Vehicle Engines," 33, No. 2, Part II, Department of Health,
Education, and Welfare (January 1968).
Federal Register, ibid, J33, No. 188 (1968).
Fleach, J. P., Norris, C. H., and Nugent, Jr., A. E., Amer Ind Hyg
ASBOC. J_. , 28, 507 (1967).
Friberg, L. , Piscator, M. , Nordberg., G. , "Cadmium in the Environment."
CRC Press, Cleveland, Ohio (1971).
Friberg, L., Piscator, M., Nordberg, G. F., and Kjellstrom, T.,
Cadmium in the Environment, 2nd edition, CRC Press, Cleveland,
Ohio (1974).
Friedlander, S. K., Smoke, Dust, and Haze: Fundamentals of Aerosol
Behavior. John Wiley and Sons, Inc., New York, New York (1976).
Fulkerson, W. and Goeller, H. E. (eds) "Cadmium: the Dissipated
Element" Oak Ridge National Lab. Report ORNL-NSL-EP-21 (1973).
Ganley, J, T., and Springer, G. S., Environ. Sci. Technol.. 8, 340
(1974).
Gillette,: D. A., and Winchester, J. W., "A Study of Aging of Lead Aerosols,"
Unll,-. of Mich, special report, Department of Meteorology and Oceanography
(February 1970).
REFERENCES (Continued)
Goetz, A., Stevenson, H. J. R., and Preining, 0., £• APCA, 17, 664
(1967).
Gregory, P. H., Microbiology of the Atmosphere, Leonard Hill Books, Ltd.,
London, 1961, pp. 63
Habibi, K. , Environ. Sci. Technol. , ^, 239 (1970).
Habibi, K., Environ. Sci. Technol. , 2. 223 (1973).
Harrison, P. R. , Matson, W. R. , and Winchester, J. W. , Atmos. Env.
_5, 613 (1971).
Heichel, G. H., and Hankin, L., Environ. Sci. Technol., £, 1121
(1972).
Bering, S. V., unpublished data for lead obtained with a modified Battelle
low pressure impactor, Caltech (1975).
Herring, W. 0., EPA Report PB-201 739, APTD-0706 (1971).
Hidy, G. M. , "Characterization of Aerosols in California," Interim
Report for Phase I, State of California Air Resources Board Contract
No. 358 (1973).
Hidy, G. M., "Characterization of Aerosols in California," Final
Report, Vol II, State of California Air Resources Board Contract
No. 358 (1974).
Hidy, G. M., Mueller, P. K., Wang, H. H., Karney, J., Twiss, S.,
Imada, M. , and Alcocer, A., J_. Appl. Meteor. , 13, 96 (1974).
Hirschler, D. A., Gilbert, L. F., Lamb F. W., and Niebylski, L. M.,
Ind. Eng. Chem., 49 1131 (1957).
Hirschler, D. A., and Gilbert, L. F., Arch. Environ. Health, 1, 297 (1964).
Huntzicker, J. J., Davidson, C. I., and Friedlander, S. K., "The Flow of
Trace Elements throughout the Los Angeles Basin: Zn, Cd, and Ni,"
Division of Environmental Chemistry, 167th Meeting, ACS, Los Angeles,
California (March 1974).
Huntzicker, J. J., and Davidson, C. I., "Atmospheric Trace Metal Flows
in the Los Angeles Basin: Zn, Cd, and Ni," in "Air-Water-Land
Relationships for Selected Pollutants in Southern California,"
Huntzicker, J. J., Friedlander, S. K., and Davidson, D. I., Ed.,
Final Report to the Rockfeller Foundation, Caltech (August 1975).
108
109
-------
REFERENCES (Continued)
Huntzicker, J. J., Frledlander, S. K., and Davidson, C. I., "A Material
Balance for Automobile Emitted Lead in the Los Angeles Basin,"
Environ. Set. Technol. ±, 448-457 (1975a).
Huntzicker, J. J., Friedlander, S. K., and Davidson, C. I., "Air-Water-
Land Relationships for Selected Pollutants in Southern California,"
Final Caltech Report to the Rockefeller Foundation (August 1975b).
Islitzer, M. F., and Dumbauld, R. K., Int. J. Air Wat. Poll.. 1, 999 (1963).
Kloke, A., and Rlebartach, K., Die Katurviasenahaften, 5_1, 367 (1964).
Lagerwerff, J. V., and Specht, A. W., Environ. Sci. Technol., .4, 583
(1970).
Lazrus, A. L., Lo range, E. , and Lodge, Jr., J. P., Environ. Scl. Technol., 4^
55 (1970).
Lee, Jr., R. E., Patterson, R. K., and Wagman, J., Environ. Sci. Technol.. 2,
288 (1968).
Lee, Jr., R. E., Patterson, R. K., Crider, W. L., and Wagman, J.,
Atmos. Env., _5, 225 Q971).
Lee, Jr., R. E., and von Lehmden, D. J., J_. APCA, 23, 853 (1973).
Lee, Jr., R. E., and Goranson, S. S., "Variations in the Size of
Airborne Particulate Matter over a Three-Year Period," 68th
Annual Meeting, AIChE, Los Angeles (November 1975).
Lees, L., "Smog—A Report to the People," Environmental Quality Laboratory,
California Institute of Technology, Pasadena, California (1972).
Leh, H. 0., Gesunde Pflanzen, 18, 21 (1966).
Lemke, E. E., "Profile of Air Pollution Control," Air Pollution Control
District, County of Los Angeles (1971).
Linlnger, R. L., Duce, R. A., Winchester, J. W., and Matson, W. R.,
J. Geophya. Res., _71 2457 (1966).
Lundgren, D. A.. J. APCA, 20, 603 (1970).
Lundgren, D. A., and Paulus, H. J., "The Mass Distribution of Large
Atmospheric Particles—and how it Relates to What a High Volume
Sampler Collects," 66th Annual Meeting, Air Pollution Control
Association, Chicago, Illinois (June 1973).
110
REFERENCES (Continued)
Lundgren, D. A., and Paulus, J. J., "Direct Determination of the Total
Atmospheric Mass Distribution," Division of Environmental Chemistry,
167th Meeting, ACS, Los Angeles, Calif. (March 1974).
Marple, V. A., and Liu, B. Y. H., Environ. Sci. Technol.. j8, 648 (1974).
Marple, V. A., Liu, B. Y. H., and Whltby, K. T., Aerosol Set. 5, 1
(1974).
Marten, G. C., and Hammond, P. B., Agron. J_., 58, 553 (1966).
May, K. R., Applied Microbiology. 12_, 37 (1964).
McCrone, W. C., and Delly, J. G., Particle Atlas, Edition 2, Volume 2,
Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan (1973).
Megaw, W. J., and Chadvick, R. C., 1957, data published by Chamberlain,
A. C. (1960).
Milde, R. L. and Beatty, H. A., in "Metal-Organic Compounds," Advances
in Chemistry Ser., No. 23, 306, ACS, Washington, D. C. 0959).
Moran, J. B., and Manary, 0. J., "Effect of Fuel Additives on the Chemical
and Physical Characteristics of Particulate Emissions in Automotive
Exhaust Interim Technical Report to the National Air Pollution
Control Administration," Dow Chemical Co., APTD 0618, PB 196
783 (1970).
Motto, H., Dalnes, R., Chilko, D., and Motto, C., Environ. Sci.
Technol., ^, 231 (1970).
Mueller, P. K., Helwlg, H. L., Alcocer, A. E., Gong, W. K., and Jones, E. E.
ASTM Special Tech. Publication No. 352, 60 (1964).
National Air Surveillance Networks, Environmental Protection Agency Report
PB 224 823 APTD 1467 (1973).
Nlnomlya, J. S., Bergman, W., and Simpson, B. B., to "Proceedings of the
Second International Clean Air Congress," H. M. Englund and
W. T. Beery, Ed., Academic Press, New York (1971).
Ondov, J. M., Zoller, V. H., and Gordon, G. E., "Trace Elements on
Aerosols from Motor Vehicles," presented at the Air Pollution
Control Association Annual Meeting, Denver. Colorado (June 1974).
Page, A. L., Ganje, T. J., and Joshi, M. S., Hilgardia. 41. 1 (1971).
Parkhurst, J. D., "Technical Report: Haste Discharge to the Ocean,"
Sanitation Districts of Los Angeles County (January 1973).
Ill
-------
REFERENCES (Continued)
Patterson, C. and Settle, D., "Contribution of Lead via Aerosol Deposition
to the Southern California Bight," Journal de Recherche Atmospheriques
(1974).
Patterson, C. C. , and Ellas, R., unpublished data, 1975.
Peirson, D. H., Cause, P. A., Salmon, L., and Cambray, R. S., Nature,
241, 252 (1973).
Peirson, D. H., 1973, published in Chamberlain, A. C. (1974).
Pierson, W. R., and Brachaczek, W. W., Rubber Chem and Tech, 47,
1275 (1974).
Pitt, R. E. and Amy, G., "Toxic Materials Analysis of Street Surface
Contaminants," Environmental Protection Agency Report EPA-R2-73-283
(1973).
Public Health Service, Division of Air Pollution, U. S. Department
of Health, Education, and Welfare, "Survey of Lead in the Atmosphere
of Three Urban Communities," PHS 999-AP-12 (1965).
Purdue, L. J., Enrlone, R. E., Thompson, R. J., and Bonfield, B. A.,
Anal. Chem. 45, 527 (1973).
Rabinowitz, M. , Chemoaphere, ±, 175 (1972).
Rains, D. W. , and Thornton, S. , "Vegetation as an Indication of Long-
Term Lead Pollution, Its Extent and Effect," University of
California, Davis, California (1970).
Rao, A. K., "An Experimental Study of Inertial Impactors," Ph.D. Thesis
under K. T. Whitby, Particle Technology Laboratory, Mechanical
Engineering Department, University of Minnesota (June 1975), pg. 100.
Raybold, R. L., Byerly, R., Jr., "Investigation of Products of Tire
Wear," NBS Report 10834 (April 1972).
Roberts, P. J. W., Roth, P. M., and Nelson, C. Li, "Contaminant
Emissions in the Los Angeles Basin," Appendix A of "Development
of a Simulation Model for Estimating Ground Level Concentrations of
Photochemical Pollutants," Report 71 SAI-6, Systems Applications, Inc.,
Beverly Hills, California (1971).
Robinson, E. , and Ludwlg, F. L. , ±. APCA, V7, 664 (1967).
Rolfe, G. L., and Haney, A., "An Ecosystem Analysis of Environmental
Contamination by Lead," University of Illinois at Urbana-Champaign,
Institute for Environmental Studies, Metals Task Force Research
Report No. 1 (August, 1975).
112
REFERENCES (Continued)
Sanitation Districts of Los Angeles County, "Technical Report: Waste
Discharge to the Ocean" (January 1973).
Schuck, E. A., and Locke, J. K., Environ. Sci. Teehnol.. ^, 324 (1970).
Sehmel, G. A., J. Geophys. Res., 75, 1966 (1970).
Sehmel, G. A., in "Assessment of Airborne Particles," T. T. Mercer,
P. E. Morrow, W. Stober, Ed., Charles C. Thomas, Springfield,
Illinois (1972).
Sehmel, G. A., J^. Aerosol Scl. . 4^ 125 (1973).
Sehmel, G. A., Sutter, S. L., and Dana, M. T., "Dry Deposition Processes,"
Battelle NW Laboratory Paper 1751 Part 1 (1973).
Servant, J., "Deposition of Atmospheric Lead Particles to Natural
Surfaces in Field Experiments," Conference of Atmosphere-Surface
Exchange Processes, Richland, Washington (September 1974).
Servant, J. , "Le Bilan Hydrometeorologique du Plomb sur la France"
(1972-3). "Considerations sur le Bilan et Importance du Depot Sec.,"
International Conference on Trace Metals in the Environment, Toronto
(October 1975).
Shelton, E. M., "Motor Gasolines, Winter 1971-72," Mineral Industry
Surveys, Petroleum Products Survey, No. 75, U. S. Department
of the Interior, Bureau of Mines (June 1972).
Southern California Coastal Water Research Project (SCCWRP), "The
Ecology of the Southern California Bight: Implications for Water
Quality Management," SCCWRP TR104 (1973).
Southern California Coastal Water Research Project, Annual Report,
El Segundo, California (1974).
Southern California Coastal Water Research Project, Annual Report,
El Segundo, California (1975).
Southern California Coastal Water Research Project, Annual Report,
El Segundo, California (1976).
State of California, "Implementation Plan for Achieving and Maintaining
the National Ambient Air Quality Standards" (January, 1972).
State of California Water Resources Control Board, "Water Quality
Control Plan for Ocean Waters of California" (July 6, 1972).
113
-------
REFERENCES (Continued)
Tepper, L. B., "Seven-City Study of Air Pollution Lead Levels: An
Interim Report," Environmental News, U.S. Environmental Protection
Agency, June 4, 1971.
Ter Haar, G., Env. Sci. and Tech.. ^, 226 (1970).
Ter Haar, G. L., Lenane, D. N., Hu, J. N., and Brandt, M., J_. Air
Pollut. Control Assoc. 22_, 39 (1972).
Tufts, B. J., Analytical Chen., 31, 238 (1959).
U.S. Department of Commerce, "Cllmatological Data, California," National
Oceanic and Atmospheric Administration, Environmental Data
Service (1968-1973).
U.S. Geological Survey, "Water Resources Data for California, Part 2,
Water Quality Records," U.S. Dept. of Interior (1973, 1974, 1975).
Watson, H. H., Amer. Ind. Hyg. ABBOC. ±., 15, 5 (1954).
Weinstock, B., and Niki, H., Science 176. 290 (1972).
Wilson, W. E., Jr., j;. Phys. Chen. Ref. Data, _!, 535 (1972).
Zlmdahl, R. L., ID "Impact on Man of Environmental Contamination
Caused by Lead," H. W. Edwards, Ed., Colorado State Univeristy,
NSF Grant GI-4 Interim Report (1972).
114
•o.s. atraaxan PRINTING OFFICE: 1978 - 786-068/1245 Region NO. 9-1
------- |